 

 

 

 

 

 

, Q‘..

‘54,? 2.“... .ﬂ.ﬁ€1.2n+ﬁlw

 

 

THEME
200x

llBRARY
MI Michigan State
University

 

 

 

This is to certify that the
thesis entitled

Breast Cancer Screening in
Three Michigan Family Practice Clinics

presented by

Suiying Huang

has been accepted towards fulﬁllment
of the requirements for

Master's degree in Epidegiology

ale

M or professor

 

0-7639 MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

 

DATEDUE I DATE DUE DATE DUE
' ' 3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p65-p. 15

 

Breast Cancer Screening in Three Michigan
Family Practice Clinics

By

Suiying Huang

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Epidemiology

2001

n:

ABSTRACT

BREAST CANCER SCREENING IN THREE MICHIGAN FAMILY PRACTICE
CLINICS

By
Suiying Huang
As part of a research project supported by the Department of Defense on training
physicians for proper follow-up of breast abnormalities, we calculated the breast
cancer (BC) screening rate for women 40-70 years old in three Michigan Family
Practice Clinics (FPC) between 5/1/98 and 7/31/99. Breast care related ofﬁce
visits and phone calls for all eligible women in the clinics were abstracted.
Symptomatic women were eliminated from the calculation. The screening rates
for CBE performed alone were 56.5%, 50.3%, and 27%. The rates for
mammography were 55.4%, 36.0%, and 28%, and 94% of women had the
mammogram done within 3 months of recommendation. The percentages of
women who had both CBE and mammography were 35.8%, 22.8%, and 16.7%.
Among them, 90% had both tests done within 3 months. For women >=50, the
mammography screening rates were consistently higher than for women < 50, for
all three clinics. CBE screening rates varied between the two age groups. These
results underline two important points: (1) the current BC screening rates for CBE
and mammography individually or combined are unacceptably low (2) when
screening is recommended, it is accomplished 90% of the time within 3 months.
To meet the Healthy People 2000 recommended mammography and CBE
combined screening rate of 60%, interventions to improve these ﬁndings at FPC

will be urgently needed.

To Mom and Dad

iii

(n

ACKNOWLEDGEMENTS

I would like to thank my thesis advisor Dr. Dorothy Pathak for her continuous
guidance and support. I am also indebted to Dr. Janet Osuch, a highly respected
breast surgeon. It is unlikely that we would have achieved much of what we did
without her insightful suggestions and expertise in the breast cancer ﬁeld. I
believe she was correct in saying “Suiying owes me big”. I would additionally like
to thank Dr. Ellen Velie for her time and stimulating discussions. The last two
years that I worked on the DOD project had been both a rewarding and
educational experience. I would like to thank all the members on this project, with
special acknowledgement of Dr. Jodi Holtrop, Dr. Barbara Givens, Maria Struck,
and all the nurses who abstracted medical charts for us. I would like to extend a
special thanks to Dr. Jianping He, for all that she has taught me during these two
years. I would further like to thank Dr. Wilfred Kannaus for his guidance, as well
as all the team members in the Fisheaters project. It has been great fun working
with all of you, and I thank everyone for your patience with me. Last, but not the

least, I would like to thank my parents for their endless love and support.

iv

 

ME

TABLES OF CONTENTS

LIST OF TABLES ...................................................................................... vii

LIST OF FIGURES .................................................................................. viii

LIST OF ABBREVIATIONS ....................................................................... ix
CHAPTER 1

INTRODUCTION ....................................................................................... 1

I. Breast Cancer Screening Evaluation .................................... 3

A. Efﬁcacy of screening ....................................................... 3

Randomized Trials ..................................................... 3

Meta-analysis ............................................................. 7

B. Effectiveness of screening .............................................. 7

C. Efﬁciency of screening .................................................... 9

Cost ........................................................................... 9

Risks .......................................................................... 9

D. Summary ....................................................................... 10

ll. Current breast cancer screening rates ............................... 15

A. Patient self-reported breast cancer screening rates ...... 15

B. Physician self-reported breast cancer screening rates.. 16

C. Chart-audited breast cancer screening rates ................ 18

D. Comparisons between self-report and chart audit ......... 22

E. Summary of breast screening literature review ............. 23

Ill. Barriers to screening .......................................................... 26

IV. Overall study objective ....................................................... 27
CHAPTER 2

METHOD ................................................................................................. 28

I. Data source ........................................................................ 28

ll. Study population ................................................................. 28

III. Data collection .................................................................... 30

IV. Quality control audit process .............................................. 30

V. Screening rate calculation .................................................. 31

VI. Statistical analysis .............................................................. 34

CHAPTER 3

RESULTS ................................................................................................ 35
I. Sample size ........................................................................ 35
ll. Breast cancer screening rates during the 15-month study
penod ................................................................................. 39
III. Time intervals between CBE and mammography .............. 40
IV. Time intervals between mammography ordering and
actually performed .............................................................. 41
V. Breast cancer rates during an annual well-woman exam 45
VI. Breast cancer screening rates among women who did not
receive an annual well-woman exam during 5/1/98 and
7/31/99 ............................................................................... 49
VII. Compliance rate ................................................................. 51
VIII. The association between the total numbers of visits and
the breast cancer screening rates ...................................... 52
CHAPTER 4
DISCUSSION ........................................................................................... 54
l. Breast cancer screening rates ............................................ 54
II. Time intervals ..................................................................... 55
III. Breast cancer screening during an annual exam ............... 55
IV. Total numbers of ofﬁce visits and the screening rates ........ 56
V. Chart audit vs. self-reported interviews .............................. 57
VI. Interventions to increase breast cancer screening rates 58
VII. Study strength .................................................................... 59
VIII. Study limitations ................................................................. 60
CHAPTER 5
CONCLUSION ......................................................................................... 61
BIBLIOGRAPHY ...................................................................................... 62
APPENDICES .......................................................................................... 68

vi

TABLE 1.

TABLE 2.

TABLE 3.

TABLE 4.

TABLE 5.
TABLE 6.

TABLE 7.

TABLE 8.

TABLE 9.

TABLE 10.

TABLE 11.

TABLE 12.

LIST OF TABLES

Summary of Randomized Clinical Trials Testing the
Effectiveness of Screening ................................................ 12

Summary of Other Studies Testing the Effectiveness of
Screening ........................................................................... 14

Literature Review For Breast Cancer Screening Rate ........ 24

Numbers And Percentages of Eligible Patients In The
Three Clinics, Broken Down By Eligibility Criteria .............. 37

Annual BC Screening Rates Among Group A Women ....... 42
Annual BC Screening Rates By Age Groups ...................... 43

Annual BC Screening Rates By Age Groups Including
Women In Group B ............................................................. 44

Screening Rates In Women Who Received A Well
Woman's Exam (WW) During 5/1/98 and 7l31/99 .............. 47

Percentages of CBE Performed And Mammography
Recommended During WW ................................................ 48

Screening Rates Among Women Who Did Not Receive
a Well Women (WW) Exam During 5/1/98 and 7/31/99 ...... 50

Reasons And Numbers of Refusals When Test Is
Recommended ................................................................... 51

Odds Ratios and 95% Conﬁdence Intervals For the

Association Between Total Number of Visits and BC
Screening Rates ................................................................. 53

vii

LIST OF FIGURES

FIGURE 1. Logistic Flow Chart ............................................................. 38

viii

LIST OF ABBREVIATIONS

BC - Breast Cancer
CBE - Clinical Breast Exam

ACS - American Cancer Society

ix

CHAPTER 1 INTRODUCTION

Breast cancer (BC) is the most common cancer among women, and it is
the second leading cause of cancer death in women, next to lung cancer. The
American Cancer Society (ACS) estimated that there will be 182,800 new cases
of invasive BC among women and about 40,800 deaths in the United States
during 2000 [1]. Based on current incidence rates, ACS estimates that onebout of
every nine women in the United States will develop BC at some time during her

life.

One effective strategy in reducing mortality from cancer is early detection
by screening. Early detection of cancer can result in treatment before the tumor
metastasizes and can lead to reduction in mortality from the disease. For a
screening test to be effective, that test must be capable of diagnosing disease

prior to it becoming symptomatic [2].

The main screening methods for BC have been mammography and
clinical breast examination (CBE) performed by trained health professionals.
Mammography can generally detect smaller tumors than those found by CBE

(1 .5cm versus 1.8cm) [3].

Recommendations for screening in normal-risk women in the US vary by

cancer research organizations. Every major professional cancer organization

recommends screening in women 50-69 at intervals of 1-2 years [3].
Recommendations are inconsistent for women aged 40-49 and 70 and over. The
American Cancer Society and the American College of Radiology recommend
annual mammography and CBE for women 40 to 49 years, while the National
Cancer Institute (NCI) recommends screening mammography every 1 to 2 years
for women of the same age group [3] [4]. All of these recommendations apply
only to asymptomatic women. The frequency and type of examination for
symptomatic and high-risk women will vary individually and should be determined

by the responsible physician.

Further, it is recognized that in order to eliminate the false negative rates
of either CBE or mammography alone, the two tests should be done as close in
time as possible [5]. Hicks et al found that the individual sensitivities of
mammography and CBE for detecting BC were 62% and 24%, respectively.

However the sensitivity of the two methods combined was 75% [5].

Historically, CBE has been a neglected part of the annual physical
examination. Many physicians attribute this to lack of adequate training of CBE in
medical school and also to the unrealistic amount of time that is required for
doing a proper exam [6]. In addition, several investigators have recently reported

that as the use of mammography increases, CBE usage has decreased [3] [7]

[8].

I. Breast Cancer Screening Evaluation

A. Efﬁcacy of screening

Efﬁcacy, as deﬁned by Last, is the extent to which a speciﬁc intervention
produces a beneﬁcial result under ideal conditions [2]. Efﬁcacy of screening can
be determined through randomized clinical trials, and there have been several

randomized clinical trials testing the value of BC screening (Table 1).

Randomized Trials

The ﬁrst of these, and the only one conducted in the US, was the Health
Insurance Plan of Greater New York (HIP) study, which began in 1963 and
ended in 1986. The primary objective of the study was "to determine whether
periodic breast cancer screening utilizing mammography and clinical examination
holds substantial promise for a long-term reduction in mortality from breast
cancer in the female population” [9] [10]. Women aged 40-64 years were
enrolled and were randomized individually. The screened group numbered
30,131, compared to a control group of 30,565. Each woman in the intervention
group was invited for an initial mammogram and three 12-month interval two-
view follow-up mammograms, plus clinical examinations. Women in the control
group followed their usual patterns of care. After 10 years, the cumulative
mortality from BC was reduced 29% (RR = 0.71, CI 0.55 - 0.92) in the study
group compared to the control group. However, the reduction in mortality differed

by woman’s age of entry to the study. For women younger than 50 years, the RR

was 0.81 (CI 0.53 — 1.24). Among women older than 50 years, the RR was 0.65

(CI 0.46 — 0.92).

Two randomized mammography screening trials were initiated in Sweden
in the mid-1970’s. The Malmo trial was initiated in 1976. Women enrolled were
aged 45-69 years. Subjects were randomized for an 18-24 months interval, one-
view, mammographic screening as part of their usual medical care. Women in
the control group did not receive screening. After 9 years of follow-up, the RR for
all women in the screened group was 0.96 (CI 0.68 — 1.35). Among women aged
50 years and older at entry, the RR was 0.79 (CI 0.51 - 1.24). Among women
aged less than 50 years at entry, the RR was 1.29 (CI 0.74 — 2.25). However, in
an analysis done in women 40-49 after 12 years of follow-up, the RR became
0.64 (CI 0.45 - 0.89) [11]. The results showed that mammograghic screening

may lead to reduced mortality from BC after long-term follow-up.

In 1977, the Swedish National Board of Health and Welfare started
another randomized controlled trial in two counties (Kopparberg and
Ostergotland counties) to determine the effect of screening with a 24-33 month
interval, one-view, mammogram on reducing mortality from BC [12]. Women in
the control group followed their usual patterns of care. With an average of 13
years of follow-up, the cumulative mortality from breast cancer was 30% lower in
the study group than it was in the control group (RR = 0.7, CI 0.55 - 0.87). The

effect of screening was almost entirely concentrated among older women. In

Kopparberg county, the RR was 0.73 (CI 0.31 - 1.4) in women < 50 years, and
0.58 (CI 0.43 - 0.78) in women older than 50. In Ostergotland county, among
women < 50 years old, the RR was 1.02 (CI 0.52 - 1.99), and for women 50

years and older, the RR was" 0.73 (CI 0.56 — 0.97).

In another randomized clinical trial conducted in Edinburgh, 46,000
women aged 45-64 years were recruited during the period of 1978-1981. The
screening methods included an annual two-view mammogram and CBE. Women
in the control group received routine health care. After 7 years of follow-up, a
non-signiﬁcant mortality reduction was observed among women < 50 years of
age at entry (RR = 0.98, CI 0.45 — 2.1 ). Among women >= 50 years at entry, the
RR was 0.80 (CI 0.54 — 1.17) [13]. In an analysis performed in women less than
50 after 12 years of follow-up, there was a non-signiﬁcant mortality reduction of

15% (RR = 0.85, CI 0.55 — 1.41) [14].

Another Swedish trial, the Stockholm trial, was initiated in 1981 [17] [18].
The number of women aged 40 to 64 in the Intervention arm was 40,000, while
the number in the control group was 20,000. The screening method used was a
one-view, 28-month interval mammography. Women in the control group did not
receive screening. After follow-up of 11.4 years, a non-signiﬁcant 26% mortality
reduction was observed in all women in the intervention group (RR = 0.74, CI 0.5

— 1.1). Beneﬁcial effects were observed in women older than 50 years (RR =

0.62, CI 0.38 — 1.0). For women aged 40-49 years, no effect on mortality was

found (RR = 1.08, CI 0.5 - 1.7).

The Canadian National Breast Screening study enrolled 90,000 women
40-59 years of age, starting from 1981. These women were randomly distributed
into an intervention group receiving both annual two-view mammography and
CBE or into a control group receiving only annual CBE [15, 16]. After 10.5 years
of follow-up, among those women aged younger than 50 at entry, the RR of
mortality from BC for those in the intervention group was 1.14 (CI 0.83 - 1.56),
compared to controls. Among women aged 50 years and above, the RR was
0.97 (CI 0.62 — 1.52). Their results showed that screening with yearly two-view
mammography and CBE had no impact on the rate of death from breast cancer

for up to 10 years of follow-up from entry in this trial.

The Gothenburg breast cancer screening trial started in 1982 in Sweden.
The trial randomized 52,000 women aged 40 - 64 into two groups: one received
mammographic screening every 18 months, and one control group, who was not
invited to screening until the ﬁfth screen of the intervention group [19] [20]. After
7 years of follow-up, no signiﬁcant reduction in mortality in all women in the
screened group was observed. However, after 12 years, there was a signiﬁcant
44% reduction in mortality from BC in the screened group of women < 50 years

at entry compared to the control group (RR = 0.56, CI 0.32 — 0.98) Their data

suggested that at least 10-12 years of follow-up is needed for the reduction in

mortality to be seen among women under the age of 50.

Meta-analysis

Hendrick et al conducted a meta-analysis of eight randomized controlled
trials of screening mammography involving women aged 40-49 at entry [21]. The
average follow-up time was 12.7 years. The meta-analysis was performed using
a Mentel-Haenszel estimator method. After combining the most recent follow-up
data, a statistically signiﬁcant 18% mortality reduction among women who were
randomized to screening mammography was observed (RR = 0.82, CI 0.71 —
0.95). This meta-analysis showed, by combining all eight randomized clinical
trials involving women younger than 50 years at entry, a statistically signiﬁcant
mortality reduction due to regular screening mammography was observed. This
analysis overcame many of the power limitations in the younger age groups that
challenged the accuracy of the previous trials, due to the lower prevalence of BC

in this age group.

B. Effectiveness of screening
Effectiveness, as deﬁned by Last, is a measurement of the extent to which
a speciﬁc intervention, when deployed in the ﬁeld in routine circumstances, does

what it is intended to do for a speciﬁed population [2].

 

 

 

WC
ins

SCI

IIOI
lilo,
0.7:

454

 

la

 

ier
Bat
en
lm

life

 

allt

C

 

One of the largest tests of BC screening effectiveness was the Breast
Cancer Detection Demonstration Project (BCDDP), sponsored by the American
Cancer Society and the National Cancer institute. Between 1973 and 1981, a
total of 283,222 women aged 35-74 years participated in the BCDDP program.
The program provided annual two-view screening mammography and CBE for
ﬁve years, in 29 centers throughout the US. This project was a screening
demonstration project that did not include a comparison group of women who did
not receive mammographic screening, and so could not measure mortality
reduction. However, after 20 years of follow-up, results showed that 50-59% of
the cancers diagnosed were stage 0 or I [22]. The results demonstrated that BC
can be detected at an earlier stage among women of all ages when screening

modalities are used.

A second large-scale non-randomized trial was initiated in the United
Kingdom in 1979 to evaluate the effectiveness of mammography and CBE in
women aged 45 to 64 years. Subjects were not individually randomized and
instead screening eligibility depended on their area of residence. Women in the
screened population (n=45,841) were offered annual physical exam and biennial
mammography for 7 years. Women in the control population (n = 127,117) were
not offered screening services. After 16 years of follow-up, breast cancer
mortality was 27% lower in the study group, compared to the control group (RR =
0.73, CI 0.63 — 0.84) [23]. There was no evidence of less beneﬁt in women aged

45-46 years at entry, the effect of screening in this age group begins to emerge

SCI

after 3-4 years. After 16 years, a 30% (RR = 0.7, CI 0.57 - 0.86) reduction is
seen in women aged 45-46 years at entry. However, this trial is subject to

criticism since it is not individually randomized. Possible confounding factors,
such as inherent risk across the counties and differences in social-economic

status, should be considered when interpreting the results.

C. Efﬁciency of screening

 

In addition to efﬁcacy and effectiveness, BC screening efﬁciencies must
also be considered. Efﬁciency, as deﬁned by Last, is the effects or end results
achieved in relation to the effort expended in terms of money, resources, and

time [2].

Cost

The cost of screening is usually measured by the cost per year of life
saved. In 1995, it was estimated that cost/year of life saved by screening
mammography ranged from $6,000 - $13,000, with a median of $8,900 [24]. In
comparison, the median cost per year of life saved in the appropriate age groups
for other interventions were: $6,000 for cholesterol, $12,000 for cervical cancer,
and $42,000 for hormone replacement therapy. This demonstrated that annual

mammography compares favorably with other public health interventions.

Risks
However, there are existing potential hazards associated with BC

screening as well, especially with mammographic screening [25]. First, if earlier

time of diagnosis doesn’t translate into a reduction in breast cancer mortality for
an individual woman, then some women are given advanced notice of a cancer
diagnosis without tangible gain [26]. This can, of course, have an adverse effect
on the quality of life. Second, mammographic screening results in exposure to
low-dose radiation, and this may induce breast cancer, especially for women with
the inherited gene for ataxia-telangiectasia [3]. Third, false positive results can
lead to unnecessary breast biopsies and anxiety [26]. These patients have to
face the ﬁnancial/emotional burden of being falsely identiﬁed as a potential
cancer patient. Finally, mammography has a false negative rate in screening
settings of 10-15% [26]. This can lead to false reassurance that cancer is absent

and mislead women and their providers.

D. Summary

Despite the potential risks involved, data from clinical trials support on
average a 30% mortality reduction in BC resulting from annual or bi-annual
mammography and CBE among asymptomatic women between the ages of 50
and 69 years [27]. A meta-analysis of the randomized trials demonstrated a 18%
reduction in BC mortality from mammography screening among asymptomatic
women between the ages of 40 and 49 years. The lower mortality reduction
demonstrated in women 40-49 as compared with women 50 and over is likely
due to lack of power to demonstrate a difference based on low prevalence of BC

in this age group, the need for longer follow-up time, and the demonstrated need

10

for shorter screening intervals in younger women, due to shorter cancer sojourn

times in this population [28].

11

 

mmo ncm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

397.320de 9 :22: ms ooodw ooodm 8-8. 82 «350
$0 Em

8978.284: NF __2_2>~ m? 08.8 898 219. 82 328
34.30321 8v

Bmgmtmta.
E43985 8A
2.08.986 8v mmo cam 82

285.22 E .22: ms 89mm 89mm 5.? -32 05.55353
$53.8”; BA , £2 53%
83-338; 8v em “22>? 8 81% 518 3-9. -E: 3.5.3928
83.8.83... BA $2 5.8%
$75828 8v 4N N22: 2 B2: 83» «we. -Ei 5.3.880.
seemsegd 8v

Bmm>mimt<
$449828 of
633.033 8v 533m

98555. «was L22: B 8; 80.5 8-3 -22 65.22
sweetened 8A mmo as 82
3.33:3 8v 9 D22: 2 895 5.8 8-9. .8? <3

9.0.5 9.0.6
3.5 @5528 est .2280 >633 baton

co :3me 2:23. its... co «.83 co m; a; a; am< 2:: 5:83

 

 

@5528 he mmmcozuoetm 2: 95mm... wit... 525.0 358351 .0 ESE—5m 3. min...

12

EQEmOEEmE 39>; u N

EamcmoEEmE EDS-N n F

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

28.3.8 85 8v

wcmmtamtmtar
8.3.8 5o 8A
cm. 8.2 m; omv 82 888

28:92 2 .22: 2 08.8 80.8 8-8 -82 552358
8. 78.3 85 8A 82 888
878.982 8v 8 L22: 3 28.8. ooodIv 8-8 -82 €35.88

9.55 nap—m
A95 @5523 m2". .9960 203m eaten

so .83ng 3:28. 322... so 8e: co m; a; a; am< 85:. 8:83

 

 

13

EqmaoEEmE 32>; n N

EquoEEmE Ems-N n F

 

 

 

 

 

 

 

14

 

F ..o
o 888 .m 88:8... mmo .8 82

92> .858 .0 088-8 2 .25. > N 8 82.8 8-8 -22 8an
.8888 k... 8-8
8.888 E... 8.88
6. 2.88 2.0 8-8 8.0 8....
8.888 N... 8v 2 N2.). > F 2 232 .88 88.. -22 a...- 8:

88:8 8... Z. 21.... 32.2922...

8:05 9.96

35.95". .550 ..O :0 A95 mEcmeum m2". _o..Eoo >535 norm...

.88 8E 9.58. .98.... .6 88$ .6 m; c. n c. u om< 2:; 8.88..

 

 

 

 

 

 

 

 

 

@5528 *0 3283.023 65 9:88. moEBw .230 80 32:an "N 03a...

 

ll. Current breast cancer screening rates

A. Patient Self-Reported BC ScreeningRates

Anderson et al described the use of breast cancer screening within the US
population in 1987 and 1992 as reported in the National Health Interview Survey
[29]. In 1987, a total of 5,052 women aged 50 years or older were interviewed
and asked whether or not they had had mammography and CBE in the past year
(Table 3). In 1992, the corresponding women interviewed were 2,709. The
percentage of women who self-reported having received a mammogram in 1987
was 16.5%. In 1992, the percentage increased to 35.3%. The percentage of
women who self-reported receiving CBE increased from 41.6% in 1987 to 46% in
1992. These ﬁgures showed that the usage of BC screening modalities inCreased

between 1987 and 1992 but that levels remained low.

Coleman et al compared annual BC screening rates from a telephone
survey conducted in 1988 and again in 1991, among women aged 65 -— 74 [8].
Participants were selected from ﬁve communities around the country. In 1988,
the numbers of women included were 57 in California, 133 in Massachusetts,
124 in North Carolina, 64 in Long Island, and 121 in Philadelphia (Table 3). In
1991, 237 women participated in California, 508 in Massachusetts, 409 in North
Carolina, 523 in Long Island, and 479 in Philadelphia. None of the eligible
women had a previous history of BC, and all were able to complete the interview
or questionnaire. The authors found that mammography use increased from 19-

33°/o in 1988 to 35-59% in 1991. However, among women who received a

15

mammogram, the percent who also received a CBE decreased from 95% to 85%
(P = 0.001 ). They conclude that even though mammography in older women

increased dramatically over the 3 years, the use of CBE may be decreasing.

The Centers for Disease Control’s 1997 Behavioral Risk Factor
Surveillance System (BRFSS) examined the usage of screening mammography,
screening CBE, and both examinations among a multistage probability sample of
women aged 50 years and older, in 52 states (including the District of Columbia
and Puerto Rico) [30]. They used a standard questionnaire to conduct random-
digit-dialing telephone surveys. The questionnaire included questions abOut CBE
and mammography. The report was restricted only to screening examinations,
which is deﬁned as an examination that was part of a routine Check-up. In 1997,
the average percentage of women aged 50 years and older who self-reported
receiving a screening mammogram in the previous two years was 73.7%;

screening CBE 77.0%; and both examinations 66.4% (Table 3).

B. Physician Self-Reported BC Screening Rates

Albanes et al conducted a survey of physicians in Pennsylvania to
ascertain current BC early detection practices in 1988 [31]. They found that over
90% of the physicians self-reported having performed annual breast physical
examinations in asymptomatic women age 50 years or older (Table 3). However,
for this age group, annual mammograms were self-reported as ordered by only

42% of physicians.

16

Kripalani et al did a survey of self-reported BC screening rates among 700
randomly chosen Texas primary care physicians in 1996, in order to determine
their screening behaviors and compliance with national recommendations [32].
For women between 40 and 49 years of age, 75.5% of physicians reported
recommending mammography every 1-2 year(s), and 8.4% suggested screening
annually (Table 3). For women 50 years and older, 81.4% reported
recommending annual mammography and 16.1% of clinicians recommended
screening every 1 to 2 years. The authors concluded that the screening practices
reported by this sample of Texas physicians compared very favorably with those

reported by other authors.

Slanetz et al conducted questionnaires among 278 physicians in the state
of Massachusetts concerning their use of BC screening in 1995 [33]. In women
aged less than 50, 144 (52%) of 278 physicians self-reported performing annual
CBE combined with screening mammography every two years, whereas 57
(21%) favored annual mammography and CBE (Table 3). In women aged 50
years and older, 232 (83%) physicians reported screening patients annually with

CBE and mammography.

17

c. Chart-Audited BC Screenigq Rates

Burns et al investigated the prevalence of CBE amOng women receiving
mammography [7]. This retrospective cohort consisted of one hundred women
aged 50 years or older who received mammography between 1987 and 1990 in
Boston, Mass. Chart review recorded demographic information, severity of
illness, and performance of CBE, within 1 year to 18 months after the
mammography. They found that 76% of the population studied had
mammography and CBE, while the remaining 24% had mammography alone.
Socioeconomic factors did not differ for women with and without screening
examinations. However, female breast care providers were more likely to perform
screening examinations (both mammography and CBE) than male providers. The
authors concluded that mammography may be replacing CBE, especially among
patients receiving breast care from male providers. Interventions that are
targeted to male providers should help to improve the use of both CBE and

mammography.

Love et al determined the frequency and determinants of mammography
screening in 24 nonacademic primary care group practices, during a 3-year
period, 1988 through 1991 [34]. They audited the medical records and obtained
questionnaire responses from 1819 women older than 50 and from their 98
physicians in the non-metropolitan Midwest. Medical record abstraction indicated
that mammography was performed in all 3 years in 16.7% of women, in at least
two of 3 years in 49.8% of women, and in at least one of 3 years in 81.7% of

women (Table 3). The signiﬁcant predictors for receiving mammography included

18

family history of BC, health insurance coverage for mammography, and greater
annual household income. The strongest predictor for greater frequency of
mammography was the discussion of the procedure by a clinic staff member. The
authors concluded that clinic staff initiatives with screening mammography have
a large impact on higher rates of mammography performed, and should be a
focus of intervention research designed to increase use of screening

mammography.

Kinsinger et al conducted a randomized controlled trial with primary care
practices to evaluate the improvement of performance rates of BC screening
through implementation of ofﬁce systems in 1992 [35]. Physicians in 20 mostly
rural counties in North Carolina were assigned to either an intervention group or
a control group. The intervention, focusing on BC screening by mammography
and CBE, consisted of a series of activities designed to assist primary care
practices in developing and implementing individualized ofﬁce systems for BC
screening. To facilitate the implementation of ofﬁce system plans in the
intervention groups, practices were encouraged to use resources for tracking and
prompting (e.g., ﬂow sheets, chart prompts and sticker, etc) and for patient
education (e.g., brochures listing recommended preventive care for women over
50 years of age). Medical records of women 50 years and older were randomly
chosen for data abstraction, both at baseline year (1992) and follow-up year
(1995). The numbers of records abstracted were 2,887 and 2,874 for the two

years, respectively (Table 3). The chart audits showed an increase from 39% to

19

51% in the mention of mammography (“mention” of mammography on the visit
note in any way) in the intervention practices, compared with increases from 41%
to 44% in the control practices (Odds Ratio = 1.5, Cl 1.1 — 2.0). However, there
was no signiﬁcant difference between the two groups in the percent of actual
mammograms reported in the charts during the two years. In the intervention
group, the percentage of women with a mammogram reported in the chart
increased from 28% to 32.7%. In the control group, it increased from 30.6% to
34.0%. Regarding CBE, either completion of CBE or mention of a CBE
recommendation was considered. The percentage of women having a CBE
either performed or recommended improved from 41.1% to 46.4% in the
intervention arm, while it dropped from 44.6% to 43.9% in the control group. The
percentages of women whose chart indicated that both mammography and CBE
were recommended increased from 28.2% to 38.7% in the intervention group,
and 30.3% to 32.6% in the control group. These results showed that outreach
interventions to increase rates of BC screening through the development of ofﬁce
systems was modestly successful in improving the documentation of
recommendation for mammography, but had little impact on the actual

performance of BC screening.

McCarthy et al measured the effect of systemic health care delivery
factors and patient demographic factors on the use of mammography among a
population of women with insurance coverage for screening mammography in

1 992 [36]. They studied 8,805 women, age >= 50 years, who were members of a

20

health maintenance organization in Michigan during 1992. Data were obtained
using computerized patient registration and billing systems. In 1992, 47% of the
entire study population received a mammogram (Table 3). Not having at least
one primary care visit at the time when due for screening was the strongest
predictor for not receiving a mammogram. This study suggested that physicians
may rely too much on offering mammography during ofﬁce visits, and that more
attention should be focused on a population-based perspective that includes
outreach to women who have not visited their health care provider and are
overdue for screening. In addition, they also found that the number of visits a
patient had was related to obtaining a mammogram. Women who had 2-10 visits
had the highest mammography use, compared to those with 1 visit and visits

beyond 10.

Tishler et al tried to determine the rates of BC screening for older women
cared for in a primary care practice in 1996 [37]. The retrospective cohort
consisted of 130 women aged 65 to 80. Data were collected from the hospital's
computerized medical record between October 1996 and October 1997. They
abstracted all CBE and mammograms performed or recommended during the 2-
year study period. They found that among the 130 women, mammography was
recommended for 95% of women and completed for 84% (Table 3). CBE was
performed on 75% of those women. They reported a very high rate of
mammography for women cared for in a hospital-based primary care practice,

about twice that reported in most previous studies. The systems in place to

21

facilitate ordering and tracking of mammograms may have contributed to the
unusually high rates of mammography observed. Mammograms were included in
a computerized “To Do” list for women aged 50 and older. The clinician received
a computer prompt at the time of a patient’s visit if it had been more than a year

since the women’s last mammogram.

D. Comparisons Between Self-report And Chart Audit

Montano et al measured the cancer screening rates of family physicians
and compared the measures obtained by physician self-reports, chart audits, and
patient surveys in 1988 [38]. Sixty physicians participated in the physician
survey, and 326 patients were surveyed for each physician (n = 21,876 patients).
Fifty to sixty patients’ charts were selected for each participating physician (n =
3,281 patient charts). The chart audit indicated that on average 51 % of female
patients older than 50 years had had a mammogram within the previous year of
the study (between 1988 and 1989), and 57% of women had had a CBE in the
past year. Corresponding physicians’ self report showed that the rate for
mammography was 51% among women aged 50 and older, and 67% for CBE.
Patients' self reported survey indicated that 46% of women older than 50

received mammography and 63% received CBE (Table 3).

Whitman et al tried to determine whether chart reviews and patient
interviews provide the same information about BC screening [39]. The

percentage of women older than 40 who received a breast exam and the

22

percentage of women older than 50 who received a mammogram at two different
public health clinics in Chicago were studied using both chart reviews and
telephone interviews. They found that interviews estimated signiﬁcantly higher
proportions of women having received breast exams and mammograms in the
previous 12-month interval than were estimated from randomly selected medical
records. At center A, the chart review produced an estimate of 6% of women who
received CBE, while patient interviews produced an estimate of 55% (Table 3).
At center B, the chart review indicated that 36% of the eligible patients had
received a CBE in the past year compared to 63% derived from the telephone
interview. Regarding mammography, 3% of the eligible patients had
mammography recorded in their charts in Center A, while interviews estimated
29%. At Center B, 17% of the women had mammograms recorded in their charts,
while interviews produced 38%. This study demonstrated that the BC screening
rates in the two clinic centers were low, and there are marked discrepancies
between what women report regarding BC screening and what is revealed by

reviewing the medical records.

E. Summary of Breast Cancer screening literature review

BC screening rates can be reported by interviewing patients, physicians,
or by medical chart auditing. Self-reported BC screening rates are consistently
higher than those rates obtained from medical chart auditing. The literature also
indicated that since the late 1980’s mammography usage had increased steadily.

However studies have reported that CBE usage may be decreasing.

23

noncoEEooo. n F

 

 

 

 

 

 

 

 

 

 

 

 

 

25:83 4. 8 om "A
LEEooEV 02 913 vmm roam. 83 i as :53 5.
.23 cm_o_m>cn_
mm mm mm om "A
Nm mm mm 24.94 who. tone. mm? _.w «o 5255
.23 cm_o_m>cn_
PA._e__..xcem..v Ne om om "A 88 :89 82 .a .e 8:82.
.tom c9063.”.
v.8 :K R om "A tone. 5? mmmmm 000
.23 ceEo>> -89
088-9.. 2.8 mm em 59
$8.? Sumo 22. tons. $9 _a «c 55200
Ema :mEo>>
m.mm 9. on "A monw Name
09 w. :V cm "A «mom tone. 5? _u .0 :oﬂeuc<
.tow cmEo>>
econ 50m e\. econ ocou em< Hz .3522 boron. >35
EEmoEEwE .\e mmo .x. :oEoB >35

 

 

 

 

 

 

 

 

 

2am 9:523 .350 “qum .5". 325m 9.39.2... .n aim-P

24

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9:5 928 new 555225 Eon *0 89805 9: E9: um>tou 92> 9mm; 92 05 E mommy—82mm .
8 SR m:
8 ow "A 8m 22225 22E
t 82 3;
on 9. "A «mm 325m :20 82 Am 5:8
3 “a $5.5,
mm 82 «3
mm ov "A 9: 32285 895
m 8% an
o ow R :3. 323m :20 82 2 .250
a 8 5.523
5 Km on "A 5% 253mm :20
9. mo om "A 89 : 32% 28$
5 my om "A can $23 cgoaén. $2 a a 88:22
vw 2 mm "A o? 295m :20 82 _~ 5 5%.
t. om "A 88 5% w :____m «2: a a 3:822
Sm v.8 «.3 on 2 E8. 225m :20 82
0mm 0mm ma... om "A $3 225m :20 $2 .3 8 8325.
a; m 5 WA Em
a; m 5 Nut 9%
a; m =5v cc 2: om "A 22 263mm. :20 -wmmmmu a .o 96..
2.8 50m .\. 0:3 2% mm< "2 3522 moron. a
Emacs—nus. .x. mmo .x. 5802. .25 L

 

 

 

 

25

Ill. Barriers to screening

Among identiﬁed barriers to screening are the discomfort or cost of the

procedure, lack of health insurance, lack of transportation or remoteness of the
mammography facility [3].

However, the two common reasons women give for not having had a
mammogram was that they did not know they needed it and that their physician

had not recommended it [40][41]. Fox et al analyzed the reasons provided by 517

women 50 years and older, living in Los Angeles, California, for their

underutilization of BC screening [40]. They found that the most important factor

that predicted whether a woman ever had a mammogram was whether her

physician had talked to her about mammography. Similar results were also found

by Grady et al [42]. Their multivariate analyses revealed physician

encouragement to be more strongly associated with screening mammography
than health status, health care utilization, attitudes, and socio-demographic

characteristics. Those women who reported having received a physician

recommendation were nearly four times more likely to have ever had a screening

mammogram than those not receiving a physician recommendation [42].

These ﬁndings further strengthened the critical importance of physician

behaviors in the secondary prevention of BC in women.

26

IV. Overall Study Objectlve

The current study was conducted to calculate the patient-speciﬁc annual
screening rates for CBE, mammography, and both, in three Michigan family
practice clinics, among women 40—70 years old. For this study, the annual
screening rate will be deﬁned as screening occurring during a ﬁfteen-month time

frame between 5/1/98 and 7l31/99.

27

CHAPTER 2 METHOD

I. Data Source:

Data for this analysis were derived from an ongoing large-scale study,
funded by the United States Department of Defense. The aim of that study was

to enhance primary care physicians’ skills in secondary prevention, diagnosis

and follow-up of abnormal ﬁndings in the control of breast cancer.

ll. Study Population:

Three mid-Michigan family practice clinics were included in this analysis.

They were designated as sites G, H, and l.

The clinics are members of the Michigan State University Network of
Family Practice Residency Programs that serve Michigan by providing family
centered care to the citizens of the communities in which they are located. They
train resident family physicians to meet primary care needs, and to reach out to
the medically underserved and the elderly of these communities. The programs

estimated that in 1996 each site saw approximately 10 to 15% of all female

patients 40 to 70 years of age. Approximately one-third of the total patients were

Medicaid patients.

28

Each site generated a list of patients who met the following criteria for

inclusion in the study:

1. Female
2. Active patients in the practice. This was deﬁned as having at least one

visit in the past three years (or since 8/1/96).

3. Between the ages of 40-70 for the baseline year, i.e. born after August 1,

1928 and before July 31, 1959

For each residency program site, two nurses with R.N degrees who were
not afﬁliated with the residency programs were recruited to conduct the audits of
the medical records. Each site was provided with one laptop computer in which
to enter and transmit data. Nurse abstractor training was held on the campus of
Michigan State University. Data entry forms were created in the ACCESS 97
database program and placed on the laptop computers. Sample cases were
identiﬁed representing a variety of breast care concerns from the Clinical Practice
Site at the Michigan State University Family Practice Center and Kalamazoo
Center for Medical Studies. Names and all identiﬁers were blacked-out.
investigators at MSU created the gold standard for the completed audits and
each of the practice cases. The nurse auditors abstracted ten sample cases and

their entries were reviewed by the investors until the abstractor achieved a

Kappa of 90% or higher as a measure of inter-rater agreement. After initial

training in August 1999, the auditors were brought back to MSU for an additional

29

day of training in September, since additional changes were made to the
database based on abstractors’ feedback. This also allowed the reinforcement of

the previously discussed audit guidelines. At the end of the training, each nurse

abstractor signed conﬁdentiality agreement forms.

ill. Data Collection

The ACCESS database (Appendix 1) captured all patient encounters and
phone calls during which breast care activities occurred. Any evidence in the
medical record of a mammogram or CBE was recorded, such as a mammogram
recommendation or report, comments regarding test refusals and comments
regarding the reasons why recommended tests were not performed. We also

recorded information regarding screening at outside facilities or by other

physicians when documented.

lV. Quality control audit process:

Two trained graduate students in Epidemiology conducted quality
assurance audits of the medical records in all three sites. The training manual
provided to the nurse abstractors was used as a reference for a one-day training
for the students. They were also required to complete the same 10 practice
cases as the nurse abstractors. These were reviewed by the investigators as

they had been for the nurse abstractors. A 100% Kappa was required from the

30

graduate students on these cases since they were to serve as the gold standard

for the abstractors.

Twelve records were randomly selected from each auditor’s list of patients
that had already been abstracted by the nurses. The complete Kappa tests for

the charts audited were shown in Appendix 2.

The “*” in Appendix 2 speciﬁes that Kappa value was 100%. Over 90% of
Kappa values were 100% and the remaining ones were either excellent (>80%)
or Very Good (GO-80%). Only 3 kappa values were less then 60% and they were
49%, 58% and 59%. This high quality of abstracting was the result of the
intensive training that the abstractors received and the requirements that for the
10 practice cases their Kappa (agreement) values be at least 90% prior to being
allowed to abstract in the ﬁeld. The additional day of training that the auditors

received prior to entering the ﬁeld also contributed.

V. Screenlng Rate Calculations

For the purpose of this analysis, the screening rate calculation is deﬁned
as screening that occurred during a ﬁfteen-month time period from 5/1/98 to
7/31/99. If a patient’s breast care was provided by other physicians such as an
OB/GYN, or if the patient was being followed by an oncologist, this was recorded
in the database, and the patient was excluded from our screening rate

calculations. Mammograms ordered for diagnostic rather than for screening

31

purposes, either on the basis of an unresolved mammographic abnormality or an
abnormal CBE, were not considered to be a screening mammogram and this
patient was also excluded from the mammography screening rate. Similarly,
patients with a diagnostic CBE, which is deﬁned as a CBE performed after
knowledge of abnormal mammogram results, were also excluded. Comments
concerning each breast care related encounter, such as refusal and the reason

why the tests were not done, were recorded and were subsequently reviewed.

For this analysis, women were classiﬁed as being “screened” if they had
received at least one CBE or Mammogram, or both within the 15-month period

between 5/1/98 and 7l31/99.

The following screening rates or issues related to screening rates were

calculated:

(1) The CBE screening rate deﬁned by an actual CBE performed in
asymptomatic women

(2) The mammography screening rate deﬁned by an actual mammogram
performed in asymptomatic women

(3) The BC (both CBE and mammography) screening rate deﬁned by both
CBE and mammography performed in asymptomatic women.

(4) The rates of CBE recommended, regardless of whether or not they were

performed.

32

(5) The rates of mammography that are ordered,’regardless of whether or not
they were performed.

(6) The time interval between performance of CBE and mammography for
asymptomatic women who had both examinations. The four time periods
chosen for evaluation were: 3 month, 3-6 months, 6-9 months, and >9 months.
(7) The time intervals between when a mammogram was ordered and when it
was actually done, according to the four intervals described above.

(8) The compliance rate for CBE and mammography: percentages of women
who refused mammography or CBE upon recommendation.

(9) The reasons for refusal if documented in charts and other reasons why
mammography or CBE was deferred or not performed.

(10) The percentages of women who received an annual well-women exam.
(11) The percentages of CBE performed and mammograms ordered during
annual well-women exams.

(12) The BC screening rate among women who did not receive an annual well
women exam.

(13) The screening rates broken down by age groups: women 40-49 and
women 50-69.

(14) The association between the total numbers of visits to the family practice
physicians during the 15-month study period and the BC screening rates. Total
numbers of visits were grouped into 1-2 visit(s), 3-4 visits, and beyond 5 visits.

Because we collected the total number of visits not only between 5/1/98 to

33

7/31/99, but also included visits that occurred before 5/1/98, the total number of

visits can only serve as a proxy indicator.

VI. Statistical Analysis

Odds ratios (OR) and 95% conﬁdence intervals (Cl), derived from logistic
regression models, were calculated to ascertain the association between the total
numbers of visits to the family practice physicians during the 15-month study

period and the BC screening rates.

34

CHAPTER 3. RESULTS

I. Sample Size

The numbers of patients assessed for eligibility in the three sites were
540, 872, and 896 (Table 4). Among them, the numbers of patients who were
ineligible for analysis were: 23 (4.3%), 94 (10.8%), and 25 (2.8%). These are the
patients who were male, not active during the last 3 years, outside the stated age
range, or whom breast care was not provided by a family practice provider
(Figure 1). The numbers of eligible women were 517 (95.7%), 778 (89.2%), and
871 (97.2%) at site G, H and I, respectively. These women presented at least
once to the ofﬁce during the last 3 years and represented the population that

should have received a CBE and mammogram.

Two BC screening rates were generated as follows:

1. BC screening rates among GROUP A women (those who had at least one
ofﬁce visit for any reason or had a phone call/reminder that’s breast
related during 8/1/98 and 7/31/99). The numbers of patients who met
those criteria were 398 (73.7%), 653 (74.9%), and 505 (56.4%), in site G,
H and l, respectively. The percentage of eligible women who were seen

between 8/1/98 and 7/31/99 and in whom no breast care was performed

35

were 87 (16.1%), 205 (23.5%) and 219 (24.5%), in Site G, H and l,

respectively (Table 4).

. BC screening rates among GROUP A and GROUP B women. GROUP B
women were those who presented at least once to the ofﬁce during the
last 3 years, but did NOT have one ofﬁce visit for any reason or have a
phone call/reminder that’s breast related during 8/1/98 and 7/31/99. The
numbers of patients under this description in the three sites were: 119
(22%), 125 (14.3%), and 366 (40.8%) at site G, H and l, respectively.
These women were included only in the denominator of our screening
rates, because they had no breast care activities during our study period

(Table 4).

Figure 1 showed details of the screening rate calculation.

36

Table 4. Numbers And Percentages of Eligible Women In The Three Clinics,

Broken Down By Eligibility Criteria

 

 

 

 

 

 

Eligible Women
Group A7 Ineligible Total
:22. 2222:: .2 w-..
Care Care
SiteG 311 (57.6%) 87 (16.1%) 119 (22.0%) 23 (4.3%) 540
Site H 448 (51.4%) 205 (23.5%) 125 (14.3%) 94 (10.8%) 872
Sitel 286 (31.9%) 219 (24.5%) 366 (40.8%) 25 (2.8%) 896

 

 

 

 

 

 

 

 

1 = Eligible women who have had one ofﬁce visit to the family practice
clinic for any reason or had a phone call/reminder that’s breast related

during 811/98 and 7l31l99

2 = Eligible women who did not have one office visit to the family practice
clinic for any reason or had a phone call/reminder that’s breast related

during 8I1I98 and 7l31l99

37

Figure 1: Logistic Flow Chart

 

 

Ineligible (patient is male;
All Potential Patients not active during the last

  
 
 

 

 

 

 

 

(Eligible and Ineligible) 3 years; age not between
40-70; breast care not
T provided by FPCP)
Group A and B (active patients in the Group 3 (not ACTIVE
last 3 years) between 8/1 I98 and

 

 

 

 

 

l 7131/99) E

Group A (ACTIVE between 811/98 and N O
—>

 
     
 

7l31l99)

   

 

    
 
  

 
     
 

 

  

 

 

 

 

 

Breast I
2 care
There is I
Breast care 1 -
f \ ﬂ. First breast care encounter \
First breast symptomatic (presenting I
care symptoms such as nipple or
encounter skin changes) I
asymptomatic _O_R
‘ ll. symptomatic after ﬁrst breast
—> cars encounter due to either I
abnormal CBE or abnormal
mammogram
\ / k j

 

 

 

 

l l l 1

Normal Abnormal Normal Abnormal
Finding Finding Finding Finding

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

Numerator Screeningj’ate Calculation

 

Denominator

 

 

 

38

II. BC screening rates during the 15-month study period

Table 5 and 6 shows the BC screening rates in women who had at least

one ofﬁce visit to the family practice clinic for any reason or had a phone

call/reminder that’s breast related during 8/1/98 and 7/31/99.

Our results shows that the percentages of CBE and mammography

conducted differed between women older than 50 years and younger than 50.

For CBE, women older than 50 had higher, lower, and equal rates at clinic
G, H and I, respectively, compared to women younger than 50. Among clinics G,
H and I, the overall percentages of women who received at least one CBE were
53.0%, 45.2%, and 27.0%, respectively (Table 5). Among women aged 40-49,

the rates were 44.0%, 49.2%, and 25.8%. Among women 50 years and older, the

rates were 59.9%, 41 .5%, 28.1% (Table 6).

For women aged 50 and older, the mammography screening rates were

consistently higher than for women younger than 50, in all three clinics. The
percentages of women who had at least one mammogram during the study
period were 52.3%, 32.5%, and 28.0%, in the three clinics, respectively (Table
5). Among women aged 40-49, the rates were 41.5%, 24.4%, and 21.7%. Among

women 50 years and older, the rates were 60.8%, 40.0%, and 34.0%(Table 6).

39

The percentages of women who had both CBE and mammogram were
35.8%, 22.8%, and 16.7%, in site G, H and l, respectively (Table 5). Among
women aged 40-49, the rates were 26%, 19%, and 14%. Among women 50

years and older, the rates were 45%, 27.3%, and 19.7% (Table 6).

Table 7 shows the BC screening rates among women in GROUP A and
GROUP B. It also demonstrate the rates in women who DID NOT have at least
one ofﬁce visit to the family practice clinic for any reason or had a phone
call/reminder that’s breast related during 8/1/98 and 7/31/99. With the inclusion of

this latter group, the screening rates were even lower (Table 7). ln site I, <10% of

all women received both CBE and mammogram.

lll. Time intervals between CBE and mammography

We examined the time interval between performance of CBE and
mammography for asymptomatic women who had both examinations. Our results
showed that in all three sites, CBE and mammography were performed within

three months of one another 90-91% of the time (Table 5).

40

IV. Time intervals between when mammography was ordered and actually

performed

We also evaluated the time interval between when a mammogram was
ordered and when it was actually done. Among women who had at least one
mammogram, 98.3%, 93.9%, and 96.2% of them had less than 3-month time
intervals between the time that mammogram was ordered and when it was

actually performed, in site G, H and l, respectively (Table 5).

41

Table 5. Annual BC Screening Rates Among Group A Women

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site G Site H Site I
CBE ordered 58.7% 54.7% 28.8%
CBE performed 53.0% 45.2% 27.0%
Mammogram ordered 63.5% 42.9% 44.2%
[Mammogram performed 52.3% 32.5% 28.0%
‘80 screening rate (within 3 month) 32.4% 20.7% 15.1%
'80 screening rate (both done any time) 35.8% 22.8% 16.7%
Both tests done within 3 month 91.0% 91.0% 90.0%
Mammogram done within 3 month of 77.3% 60.2% 56.2%
recommendation
Mammogram done anytime after 78.6% 64.1% 58.4%
recommendation
Mammogram done within 3 month of 98.3% 93.9% 96.2%

recommendation

 

 

 

 

42

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8m 8.. 5. m5 5. 3. 9m 8% New 4%
__e 998T: .58
8... mew mew m8 8.. 9m 8... RN .5 .5553
99:3": .58.
#3. «he. and 23.2 $3.. <38 95.23
.88 .255 $0.9. $38 .88 $3.. 88
some seem $3.9. 30.8 .23. 2.95
.28 .38 .3: .89. 48.8 80.3. 13
$0.8 23.8 .38 $0.8 $0.8 8.8 3
ﬂ 8A 8.9. Eek 8A 8.8 We 92 levee
_mEm $5 35

 

 

 

 

330.5 02 >m 92mm @5598 um .925 6 win...

43

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mam mmv 5v Kw mmv mmw ovm mam Nvm Amucwzam
=a 9.953 n c :33.
Km m3. omv wk mam mwd Cm owm 5m 5 96.5
new < 985 E 5:53
9.253 n : _EOP
«a... $2 «.98 «HS $03 $3.
$93 «1.3 .\oo.mm $1: $0.5. $0.8
$m.mm ﬁwdr $09. $08 $03 o\.m.wm
.hvdw $0.3 .\.m.om 0&me $99. .xRNm wcou mmu
ﬂaw: «ONE £53 £59. $5.0m $0.5m 322:0 mmu
Each 8A 3.3. Each 8A 913. Each 3A 1&1?
ﬁrm. :m% mem

 

 

 

m 3.0.5 :_ 22:25
9:63.05 3:20 om< >m £53. @5520...“ on .353 K 03.2.

44

V. BC screening rates during an annual well-woman exam

The percentages of women in GROUP A who received an annual well-
women exam were 58.0%, 43.5%, and 20.7% in site G, H, and I, respectively,
during the period of 5/1 I98 and 7/31/99. Among women 40—49, the percentages
were 52.0%, 47.6%, and 18.1%. Among women 50 years and older, the

percentages were 62.7%, 39.7%, and 20.7% (Table 8).

Table 8 shows the screening rates for women who received a well woman
exam. Among women 40-49 years old, the percentages received CBE during a
well woman exam were 76.7%, 95.9%, and 87.2%, in the three clinics
respectively. For women 50 years and older, the percentages were 83%, 93.1%,
and 76.5%. Women 50 years and older consistently received more frequent
recommendations for mammography during a well woman exam than those
younger than 50. The percentages were 63.5%, 51.9%, and 73.9%, for women

40-49 years old. Among women 50 years and older, the rates were 85.7%, 84%,
/

and 91%.

Table 9 demonstrated that of all of the CBE performed during the study
period, most were done during a well woman exam. In site G, among women
aged 40-49, 93.2% of CBE was done during a well woman exam; among women
age 50 years and older, 90% were done during a well woman exam. In ‘site H, for

women aged 40-49, 92.8% of CBE was done during an annual exam, and 89%

45

for women 50 years and older. In site I, among women aged 40-49, 62.1% of

CBE were done during a well woman exam, and 57.4% for women aged 50

years and older. Table 9 further illustrates the percentages of mammograms that
were recommended during an annual well-women exam. In site G, for women
aged 40-49, the percentage of mammograms that were recommended during a
well woman exam was 63.5%, and for women 50 years and older, the

percentage was 75%. In site H, among women aged 40-49, the percentage was
63.9%, and for women 50 years and older, the percentage was 56%. In site I, for

women aged 40-49, the percentage was 40%, for women 50 years and older, the

percentage was 30%.

46

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ mtm _

 

 

 

 

 

 

 

 

_ 0 Fri _

Il..$3~ $3: $5mmn1ll$ose 3.1.3... $on .582...
< .505 macs...
25> a on: 5:33 ho .x.
E t. E E x: 8 c .88
.83 $.33 $08 $.38 1.3...» $08
I IA, I 4.... , ......X I #3 . A -91...- -
IL!
1 mEm

 

8:”: 25 83m 8258 .33. E86 «.5825
:25 < .3230”. oc>> =oEo>> :. monom— m:_:oo._ow no min...

47

 

 

:_ BOEGULO EGLOOEENE

 

 

 

 

 

 

 

 

 

 

 

 

 

A F «2 F E :8 mm _ oh
mm 2. «N. E N: do Ejs 2.8 mm
ﬂax Ilavollv ISA 3.8 O? 3.8
_ mtw Em o mEm
>25 95.5 coocoﬁﬁooom

EnEmoEEws. u=< negated mmo ho 3.33528 "a win...

48

VI. BC screening rate among women whoﬂd not receive an annual well-

woman exam between 5I1I98 and 7l31l99

Among women who did not receive an annual exam during our study
period, the percentages of women who received CBE (during ofﬁce visits for
other medical reasons) were 6.0%, 6.8%, 11.8% for women 40-49 years, in

clinics G, H, and I, respectively. For women 50 years and older, the percentages

were 15.5%, 8%, and 14.9% (Table 10).

The percentages of mammograms ordered in patients who were not seen
for annual well-women exams (but during other ofﬁce visits, or as a result of
phone or card reminders) were 37.3%, 24.1%, and 24.1% for women 40-49
years, in clinics G, H and I, respectively. Among women 50 years and older, the

rates of mammography recommendation were 46.4%, 36%, and 48.7% (Table

10).

49

Table 10: Screening Rates Among Women Who
Did Not Receive a Well Women (WW) Exam
During 5I1I98 and 7l31l99

otal n of women

otal n of women

 

50

Vll. Compliance rate

Only 0.5-2% of women who had a CBE recommended refused the
examination at the time of the ofﬁce visit. The refusal rates for recommended
mammography were 0.8-2% at the time of recommendation by the family
practice physician. Table 11 lists the various reasons and total number of

patients who refused, if they were recorded in the medical charts.

Table 11: Reasons And Numbers of Refusals When Test ls

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Recommended
Reasons Numbers
Site G Refusal with no expLanation 3
Refused mammogram because it’s too painful 1
Total 4
Site H Due to insurance 4
Refusal with no explanation 5
Cited physician time restraint 3
CBE deferred due to menstruating 1
CBE deferred due to medical reasons / post 1
surgical braces
Total 14
Site I Due to insurance 2
Refusal with no explanation 1
Total 3

 

 

 

 

 

51

VIII. The associaim between the total numbers of visits and the BC

screening rates

The association between the total numbers of visits during the 15-month
period (proxy indicator), prior to the last ofﬁce visit during 8/1/98 and 7/31/99, and
the BC screening rates was also analyzed. We made the assumption that each
ofﬁce visit represented an equal and independent opportunity for a CBE, and
each ofﬁce visit/phone call consultation represented an equal and independent
opportunity for a mammography referral. Therefore, the likelihood of obtaining a

CBE or mammogram should increase predictably with each additional visit.

In Site G, the total numbers of visits among ACTIVE patients ranged from
1 to 28. In site H, the numbers ranged from 1 to 29. In site I, the number ranged
from 1 to 34. Logistic regression was used to analyze the association between
the total visits and the BC screening rates. Table 12 shows that in all three sites,
mammography screening rates were signiﬁcantly higher for those with beyond 5
visits, compared to those with 1-2 visit(s). In site I, CBE ordering and performed
were signiﬁcantly higher for those patients with beyond 3 visits than those with 1-
2 visit(s). In addition, the screening rates of BC (both CBE and mammography) in

site I were higher for those with beyond 5 visits.

52

Table 12. Odds Ratios and 95% Confidence Intervals For the
Association Between Total Number of Visits and BC Screening

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Rates
Site G N CBE CBE Mammo— Mammo— Both
ordered done gram gram done
ordered done
Total 53 1.0 1.0 1.0 1.0 1.0
visit: 1 -2
Total 84 0.96 1.13 1.53 1.52 1.18
visit: 3-4 (0.47- (0.56- (0.76- (0.76- (0.55-
1 .95) 2.62) 3.07) 3.06) 2.55)
Total 198 0.92 0.94 2.21 2.25 1.56
visit: (0.49- (0.51- (1 .19- (1.21- (0.79-
beyond 5 1.72) 1.73) 4.1 1) 4.18) 3.06)
Site H CBE CBE Mammo- Mammo- Both
ordered done gram gram done
ordered done
Total 101 1.0 1.0 1.0 1.0 1.0
visit: 1 -2
Total 120 1.12 1.25 1.12 1.2 0.97
visit: 34 (0.65- (0.74- (0.65- (0.66- (0.49-
1.91) 2.13) 1.91) 2.19) 1.91)
Total 318 1.13 1.2 1.29 1.99 1.33
visit: (0.72- (0.77- (0.82- (1.2- (0.76-
beyond 5 1.78) 1.89) 2.03) 3.3) 2.34)
Site I CBE CBE Mammo- Mammo- Both
ordered done gram gram done
ordered done
Total 107 1.0 1.0 1.0 1.0 1.0
visit: 1-2
Total 88 2.13 1.92 4.41 2.24 3.95
visit: 34 (1 .05- (0.92- (2.26- (0.93- (1 .04-
4.33) 4.0) 8.59) 5.4) 15.1)
Total 287 2.56 2.54 5.33 5.3 7.14
visit: (1.42- (1 .39- (3.02- (2.57- (2.18-
beyond 5 4.6) 4.63) 9.4) 11.0) 23.4)

 

53

 

CHAPTER 4 DISCUSSION

I. BC screening rates

Our results showed that 25.8 - 59.9% of women in the three clinics
received CBE, 21.7 — 60.8% received mammography and 13.7 — 45.1% received
both CBE and mammography during our study period. These screening rates are

far short of the Healthy People 2000’s recommended mammography and CBE

combined screening rate of 60%.

In addition, we found that in all three clinics, the mammography screening
rates were consistently higher among women 50 years or old, compared to those
less than 50. This seemed to be consistent with the current mammography
screening guidelines: every major professional organization recommends
mammographic screening in women 50-69 at intervals of 1-2 years [3]. However,

recommendations are inconsistent for women aged 40-49 and 70 and over.

CBE screening rates varied by site. In Site H, screening rates for CBE
were higher among women younger than 50 than those greater than 50, while in

Site G, the reverse was true. In site I, women less than 50 and greater than 50

had the same CBE screening rates.

II. Time Intervals

Our results showed that over 90% of mammograms and CBEs were done
within 3 months. The same applied to the time interval between when

mammography was ordered and when it was actually performed.

The potential explanation for why most mammograms were performed
within three months was that the impact of a physician’s recommendation was
most likely to be the strongest close to the time it is made. Longer intervals

between the time the test was recommended and actually performed may have

diluted the motivation inspired by the physician’s recommendation.

III. BC screening during an annual exam

Consistent with Conry’s results [43], we found that the percentages of
CBEs performed during an annual exam were very high in all three sites. The
percentages of women who received mammography recommendations from the
family practice physicians were also high during a well woman exam. These
results can be confirmed by the fact that extremely low percentages of women
with no well woman visit received CBE during our study period. At least in two

sites (site G and site H), over 90% of CBE was performed during an annual well-

women exam.

55

The percentages of women with no annual exam who received a

mammography recommendation were high. This may reﬂect the fact that

mammograms can be ordered by phone or mammogram reminders, in addition

to ofﬁce visit.

However, we also showed that percentages of women who received an

annual well-women exam during our study period are relatively low in all three

sites (18.1 — 62.7%). Interventions should be carried out to improve physician

and patients’ education about the importance of a well woman exam.

lV. Total numbers of ofﬁce visits and the screening rates

Our results demonstrated that the total number of visits made by a woman
during the 15-month period is related to higher screening rates. We found that

among all three sites, the mammography performed rates were higher for women

with beyond 5 visits, as compared to those with only 1-2 visit(s). In site I, the CBE

ordered and performed rates were also higher for this group, as compared to

those with only 1-2 visits.

McCarthy et al also found that the mammography rate was related to the
number of visits a patient had [36]. Women who had 2-10 visits had the highest
mammography use, compared to those with 1 visit or with visits beyond 10.

Among women with more than 10 visits, the rate is lower probably due to the fact

56

that these patients have other severe and more pressing chronic illnesses that

focus attention away from preventive health measures.

However, other investigators found that total numbers of visits are not

related to the screening rate [31].

V. Chart audit vs. self-reported interviews

It has generally been observed that there may be substantial differences
between information obtained from medical records audits and that obtained from
patient self-reported interviews. Whitman et al tried to determine whether chart
reviews and interviews provide the same information about breast cancer
screening [39]. They collected the percentage of women older than 40 who
received a breast exam, and the percentages of women aged older than 50 who
received a mammogram at two different public health clinics in Chicago. They
used both chart reviews and telephone interviews of women participants. They
found that interviews signiﬁcantly estimated higher proportions of women
receiving breast exam and mammograms in the previous 12 months interval than
were estimated from randomly selected medical records. There are several
possible reasons for the discrepancies: ﬁrst the medical records may be
incomplete; second the women being interviewed may incorrectly recall the time
when the test was performed, or even which test they obtained; third women

could be recalling tests they have done outside the clinics. Their results

57

suggested that precautions should be taken on the usage of survey data as

measures of actual performance. It should be accompanied by comparing these

measures with data of actual performance at the medical record level.

VI. Interventions to increase BC screening rates

One strategy for increasing BC screening rates is to enhance physician
referrals. A physician’s recommendation is one of the most important predictors
that a woman will receive a screening mammogram. A better understanding of
the factors that inﬂuence physician’s referral behavior is critical in designing
strategies to increase population coverage of BC screening. Enhancing

mammography referrals from primary care physicians is of particular public

health importance because they see a broad demographic and geographic
spectrum of women. Physicians’ screening mammography referral rates have
been found to vary by physician age, gender, and knowledge or attitudes.
Compared with older physicians, younger physicians have a greater tendency to
incorporate preventive care into their practice, to disagree less with evidence-

based guidelines, and to favor a more frequent screening interval for BC

screening [32] [33] [34] [35] [44].

Fletcher et al tested whether a community-wide intervention could
increase the usage of mammography screening for BC [45]. They conducted a

controlled study from 1/87 to 1/90 in two Eastern North Carolina communities.

58

During 1989, interventions were developed and aimed at primary care physician
and community participating women. Physicians underwent training sessions

about CBE skills. To reach community women, they used local media and

organizations. They also reviewed medical charts to determine the percentage of
women the physicians had referred for mammography. They found that the
percentage of women who reported receiving a mammogram increased from 35
to 55% in the experimental community and from 30 to 40% in the control
community. The intention to get a mammogram among eligible women was also
signiﬁcantly increased. Physician reports and medical record reviews in the

communities showed similar increases in the number of mammograms ordered.

VII. Study strengths

One strength of this study was that we abstracted medical records to
calculate BC screening rates in the three Michigan clinics. Summary sheets were
made for all breast care related visits that were recorded and reviewed manually.
CBE or mammography performed for diagnostic, rather than screening, purposes
were identiﬁed and excluded. Our sample sizes for clinics G, H and I were 540,
872, and 896, respectively. In addition, we performed a very comprehensive BC
screening rate calculation, including the ordered and performed rates of CBE and
mammography alone or combined, time interval between CBE and
mammography, time interval between when a mammography was ordered and

when it was performed, compliance rates for CBE or mammography after

59

recommendation, BC screening rates during well-woman exams. In addition, the

screening rates were broken down to women 40-49 and women 50-69, in order

to reﬂect the different national guidelines for the two age groups.

VIII. Study limitations

In interpreting results from the analysis, some limitations should be
considered. First, some CBE or mammography recommendations may have
been performed or verbal without being documented in the medical record.
Second, the chart audit may not be 100% reliable due to missing information. For

example, mammograms could have been performed elsewhere and not

documented in the charts.

In addition, though not a limitation of the screening rate calculations
reported, it would have been more helpful if we had collected some other
potential screening rate predictors, such as social economic status and insurance

coverage for all patients in the different clinics. These variables might contribute

to the differences in the screening rates among different sites.

60

CHAPTER 5 CONCLUSION

Our results underline two important points: (1) the current BC screening
rates for CBE and mammography individually or combined are unacceptably low
in the three family practice clinics we studied and (2) when screening is
recommended, compliance with the recommendation is above 98% and
accomplished 90% of the time within 3 months. To meet the Healthy People 2000
recommended mammography and CBE combined screening rate of 60%,

interventions to improve these ﬁndings at family practice clinics is urgently

needed.

61

BIBLIOGRAPHY:

Society, A.C., Breast Cancer facts and ﬁgures. American Cancer Soceity,

2000.
Last J., A Dictionary of Epidemiology. 1994: New York, NY: Oxford

University Press, 1995.
Oven'noyer, B., Breast cancer screening. Med Clin North Am, 1999. 83(6):

p. 1443-66, vi-vii.
Marwick, C., Final mammography recommendation ? Jama, 1997. 277(1 5):

p.1181.

Hicks, M.J., et al., Sensitivity of mammography and physical examination
of the breast for detecting breast cancer. Jama, 1979. 242(19): p. 2080-3.
Weinberger, M., et al., Physician-related barriers to breast cancer

screening in older women. J Gerontol, 1992. 47 Spec No: p. 111-7.
Burns, RB, et al., As mammography use increases, are some providers

omitting clinical breast examination? Arch Intern Med, 1996. 156(7): p.

741 -4.
Coleman, EA. and E.J. Feuer, Breast cancer screening among women

from 65 to 74 years of age in 1987— 88 and 1991. NC! Breast Cancer
Screening Consortium [published erratum appears in Ann Intern Med

1993 May 15;118(10):828] [see comments]. Ann Intern Med, 1992.

117(1 1): p. 961-6.

62

10.

11.

12.

13.

14.

15.

16.

17.

Shapiro, S., et al., Ten- to fourteen-year effect of screening on breast
cancer mortality. J Natl Cancer Inst, 1982. 69(2): p. 349-55.

Shapiro, 8., Determining the efficacy of breast cancer screening. Cancer,
1989. 63(10): p. 1873-80.

Andersson, I. and L. Janzon, Reduced breast cancer mortality in women
under age 50: updated results from the Malmo Mammographic Screening
Program. J Natl Cancer Inst Monogr, 1997.22: p. 63-7.

Tabar, L., et al., Efﬁcacy of breast cancer screening by age. New results
from the Swedish Two-County Trial. Cancer, 1995. 75(10): p. 2507-17.
Roberts, M.M., et al., Edinburgh trial of screening for breast cancer:
mortality at seven years. Lancet, 1990. 335(8684): p. 241 -6.

Alexander, F .E., The Edinburgh Randomized Trial of Breast Cancer
Screening. J Natl Cancer Inst Monogr, 1997.22: p. 31-5.

Miller, A.B., et al., Canadian National Breast Screening Study: 2. Breast
cancer detection and death rates among women aged 50 to 59 years.
Cmaj, 1992. 147(10): p. 1477-88.

Miller, A.B., et al., Canadian National Breast Screening Study: 1. Breast
cancer detection and death rates among women aged 40 to 49 years.

Cmaj, 1992. 147(10): p. 1459-76.

'Frisell, J. and E. Lidbrink, The Stockholm Mammographic Screening Trial:

Risks and beneﬁts in age group 40-49 years. J Natl Cancer Inst Monogr,
1997. 22: p. 49-51.

63

18.

19.

20.

21.

22.

23.

24.

25.
26.

Frisell, J., et al., Followup after 11 years—update of mortality results in the
Stockholm mammographic screening trial. Breast Cancer Res Treat,
1997. 45(3): p. 263-70.

Bjurstam, N., et al., The Gothenburg Breast Cancer Screening Trial:
preliminary results on breast cancer mortality for women aged 39-49. J
Natl Cancer Inst Monogr, 1997. 22: p. 53-5.

Bjurstam, N., et al., The Gothenburg breast screening trial: first results on
mortality, incidence, and mode of detection for women ages 39-49 years
at randomization. Cancer, 1997. 80(11): p. 2091-9.

Hendrick, R.E., et al., Beneﬁt of screening mammography in women aged
40-49: a new meta- analysis of randomized controlled trials. J Natl Cancer
Inst Monogr, 1997. 22: p. 87-92.

Smart, OR, at al., Twenty-year follow-up of the breast cancers diagnosed
during the Breast Cancer Detection Demonstration Project. CA Cancer J
Clin, 1997. 47(3): p. 134-49.

16-year mortality from breast cancer in the UK Trial of Early Detection of
Breast Cancer. Lancet, 1999. 353(91 68): p. 1 909-14.

Feig, S.A., Mammographic screening of women aged 40—49 years.
Beneﬁt, risk, and cost considerations. Cancer, 1995. 76(10 Suppl): p.
2097-1 06.

Jatoi, I., Breast cancer screening. Am J Surg, 1999. 177(6): p. 518-24.
Huney, SF. and J.M. Kaldor, The beneﬁts and risks of mammographic

screening for breast cancer. Epidemiol Rev, 1992. 14: p. 101-30.

27.

28.

29.

30.

31.

32.

33.

35.

Shapiro, 8., Screening: assessment of current studies. Cancer, 1994. 74(1
Suppl): p. 231-8.

Feig, S.A., Estimation of currently attainable beneﬁt from mammographic
screening of women aged 40-49 years. Cancer, 1995. 75(10): p. 2412-9.
Anderson, L.M. and 0.8. May, Has the use of cervical, breast, and
colorectal cancer screening increased in the United States? Am J Public
Health, 1995. 85(6): p. 840-2.

Bolen, J.C., et al., State-speciﬁc prevalence of selected health behaviors,
by race and ethnicity—Behavioral Risk Factor Surveillance System, 1 99 7.
Mor Mortal Wkly Rep CDC Surveill Summ, 2000. 49(2): p. 1-60.

Albanes, D., et al., A survey of physicians' breast cancer early detection
practices. Prev Med, 1988. 17(5): p. 643-52.

Kripalani, 8., AD. Weinberg, and HP. Cooper, Screening for breast and
prostate cancer: a survey of Texas primary care physicians. Tex Med,
1996. 92(12): p. 59-67.

Slanetz, P.J., et al., Screening mammography: effect of national
guidelines on current physician practice. Radiology, 1997. 203(2): p. 335-
8.

Love, R.R., at al., Frequency and determinants of screening for breast
cancer in primary care group practice. Arch Intern Med, 1993. 153(18): p.
2113-7.

Kinsinger, L.S., et al., Using an ofﬁce system intervention to increase

breast cancer screening. J Gen Intern Med, 1998. 13(8): p. 507-14.

65

36.

37.

38.

39.

40.

41.

42.

43.

44.

McCarthy, 8.0., at al., Screening mammography use: the importance of a
population perspective. Am J Prev Med, 1996. 12(2): p. 91-5.

Tishler, J., et al., Breast cancer screening for older women in a primary
care practice. J Am Geriatr Soc, 2000. 48(8): p. 961 -6.

Montana, DE. and W.R. Phillips, Cancer screening by primary care
physicians: a comparison of rates obtained from physician self-report,
patient survey, and chart audit. Am J Public Health, 1995. 85(6): p. 795-
800.

Whitman, S., et al., Do chart reviews and interviews provide the same
information about breast and cervical cancer screening? Int J Epidemiol,
1993. 22(3): p. 393-7.

Fox, S.A., P.J. Murata, and J.A. Stein, The impact of physician compliance
on screening mammography for older women. Arch lntem Med, 1991.
151(1): p. 50-6.

Fox, S.A., A.L. Siu, and J.A. Stein, The importance of physician
communication on breast cancer screening of older women. Arch Intern
Med, 1994. 154(18): p. 2058-68.

Grady, K.E., et al., The importance of physician encouragement in breast
cancer screening of older women. Prev Med, 1992. 21(6): p. 766-80.
Conry, C.M., et al., Factors inﬂuencing mammogram ordering at the time
of the office visit. J Fam Pract, 1993. 37(4): p. 356-60.

Roetzheim, R.G., S.A. Fox, and B. Leake, Physician-reported

determinants of screening mammography in older women: the impact of

66

45.

physician and practice characteristics. J Am Geriatr Soc, 1995. 43(12): p.
1398-402.

Fletcher, SW, at al., Increasing mammography utilization: a controlled

study. J Natl Cancer Inst, 1993. 85(2): p. 112-20.

67

Appendices

Appendix 1

Form I- Front-End Form

 

    
    

 
  

' Paﬂent Nam (LaSt): P .. r r r tBSting 94 east”. -.II 1. hr Ii r‘
(Hm I Data
“" mm" “mm“ Medical Record Number: -
F Add New Patient Date Of 3m; " ,Trenﬂsr,
' Abstractor‘s ID:

 

 

 

Eligibility Criteria:Check One Item For Each Statement (1-5)

1. mm gender Is: J IMeanlng of Eligibility Code:
3- Patient “MY '5 ”em “”9““ 1. I: Eligible for abstract and insertion

1928 and JUIY 1. 1959 2: Eligible for insertion only
4. Breast health care provided by

 

 

 

 

 

    

 

     

3= Ineligible
5. Active patient between 8/1/98-7/31/99
For site number 6-9:
V . _ I: Eligible for abstract
Click to Determine Eligibility Code: 2 or 3= Ineligible

 

 

 

Rules for Assigning Study ID:

 

Study 10 is a 6-digit number. The ﬁrst digit is your site number. The second digit is the Eligibility
code shown in the box above. The rest four digits are consecutive numbers starting 0001.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

      

 

 

 

n.... ....,........m. “WIN; .Tosavfspates ,,....-_1Il./.1.1j/11,

For your reference, please look in .me box on the right, ﬁnd For eligibility code = 110267
Ermmfafniﬂﬁaﬁmm” “$523313... giﬁiﬂlﬁﬁ: 33:33
I meil
Chart Review Form(0nly For Eligible Patient) ' Study "a. 7 "1 ' “

1. Date of Most RmtOﬂ'lce Visit (MHIDDIYY): I _

2. Autocalculated Date For the Last Eligibile Visit Within the Last 15 months (MM/DD/YY):
3. Total Number ofVlsite Within 15 Months, Including The Most Recent Visit: 1...... - __._ .

4.WesABreeetCerePerformedDurlngAnyof11leVieltszttlln111e 15MonthsPerlod:

7O

,I Add New Patient El
1

L.

5. Personal/Family History 01' Breast Cancer?

 

i- - - -- -- 1 [Rule for iiiiing in the age at diagnosis: I

1) Fill in exact age when information is availabe;

2) Fill In '777' if only known Pre-menopausal equal to or less than 50 years ol
3) Fill in '888' if only known Post-menopausal or greater than 50 years old;
4) Fill in '999' if no information is available.

In Self? INC 7 I . Age: I
Surgery/Reconstruction:
E] Complete Breast Removal [:| Partial Breast Removal/Lumpectomy
E] Prophylactic Implants [:1 Autologous Reconstitution
El Other, specify ‘ ‘
El Undocumented
Treatments (ched< all that apply)
[3 Chemotherapy D Radiation E] Tamoxifen/Nolvadex
[:J Altemative medidne(s), specify ‘ ’ ’ '

 

 

 

 

 

 

 

 

 

 

 

 

 

[:1 Other, specify .

E] Undocumented
InMother? Iiio ME ,,
In Sister? No . [:1 SIsterl Age: [:1 Sisterz Age:
In Daughter? No [:1 Daughterl Age: [:1 DaughterZ Age:
Inomerneiatives? INo . f) .Pleasespecify: ' ” H ‘

 

 

[BOX-A Record information for patients eadi visit when a breast care was
performed. Start with the ﬁrst visit when any breast care activity was
recorded during that 15 months period. Clidt the button on the right to
continue.

 

 

 

 

 

    

Go To ‘First litatientw Go To'Previous Patient {"50 To Next'P‘atient ‘1 rrGo ToLastl'Pati‘ent
(Clldt Any of the Buttons Above to Navigate the Record)

 

            

71

Form II- Visit Entry

'I‘." "ll Li .I. .Hi

Go To Next Visit

      

   

   

‘II'III‘. MTmUlI-litﬂ'lillﬂldlthi Iii-W ’I

réo To First Visit n] Go To“ Previous Visit

    

Ill '.ii' .‘I. L.‘ .." l-
. Go To Last Visit]
l 'i'h‘i ‘l‘.‘l'rlL‘-"3d.‘

  

 

I

Add New Visit ‘

I“ . . I_'=_F:1:L‘ and-1391:“

    

is

F11 hm ml entity.

_ . I Go Back to "east-é";
Study ID: 1
Please fill out Question 6 and Question 7 for every visit/call.

 

 

6. Date of Breast Care Activity Was Recorded: 11/11/11 If this visit is about a test
result, you can directly go to

Tvne. nf Cnnfact: ._ _ _ .
. Test Result Form, without
‘ ﬁlling out CBE documentation

I?
1'
II

7. Purpose of this Visit/Call: ‘
Specify: Co Directly to" '
Test Result Form
—
8.WlioPeiformed BreastCere/PhoneConsultadon? (CheckAllThatApply)
E] Resident Physician C] Faculty Physician El Physician Assistant El Nurse Practtioner

9. Pedant Presenting Symptoms/Signs (Check All That Apply)

   

 

Whlchbreest(s)haspresentlngsymptom? . . .
Ifyoudontknowwhldibmasbpleasencordlnfonnaﬂonln'LeRBnast'cetegow.
LettBreau: RIghtBreast:

13 None CI Undocumented/Don't know

 

 

1:] None D Undocumented/Don't know
[3 LumP(s)/MaS(s)/Asvmmetrial midtering
Dmppleobd‘arge . ,. ,1 . a
E Skin/Nipple change (cheat all that apply)
III Skin Dimpling E] Erythema/Sldn thidtenlng
CI Nipple Retraction El Nipple Scaling

CI Lump(s)/Mass(es)/Asymmetrical thidcening
[:1 NippIeDisdiarge '1 7 1' I
52] Skin/Nipple change (meet all thatappiy)

CI Skin Dimpling E] Erythema/Skin thidcening
[:I Nipple Retraction CI Nipple Scaling

 

 

C] Pain/Tenderness 4‘ . A, , [j Pain/Tenderness
I2 Occult Mammographic Abnormality E Occult Mammographic Abnormality
C] Densi‘ ‘ ‘ty(Nodule‘or Mmmetry) [j Density(NoduleorAsymmetry) A ' I 7
CI Microcaldﬁcations . CI Microcalciﬁcations
D Other: M“ T“ gj';1;;._‘ii"1;,1;_'i '3 WWW“ I _ I

 

 

10.CBE Documentation: .
11. CBE Findings (Check AllThatApply):

D Bilateral Implants

8 Previous abnormallty resolved
[:1 Lump/mass resolved C] Observational ﬁnding resolved C] Nipple discharge resolved [:1 Pain gone

[:1 Normal] Symmetrical nodularlty/ Symmetrical f‘lbrocystic(FlII Out Quality of CBE Documentation)
Quality of Written Description of CBE Documentation (Check All That Apply):

72

 

. NI e Cha Undocumented Breast Size Sha Undocumented
[:1 Inspection, speafy: pp' nge a /
Scar Undocumented Skin Change Undocumented
[:1 Palpatlon, specify: Fibrocystic Bre Undocumented Nodularity Undocumented
Ma$(es) Undocumented , Pain/tendemess Undocumented

 

[:1 Lymph node examination Adenopathy/Aidllary Nodes Undocumented

 

DNoepedflcdocumentatlonbeeldeenonnal
Dome.5ped~: ‘ ’

 

 

 

Abnormal: Which breath) has abnormal finding?

Ifyoudon'thiowwhidibreastpleaunoordinfomaﬂmln'mamast'catepow.

Lel't Breast: Right Breast:

 

Location: Location:

73

 

 

 

I] Lump(s)/Mass(es)/Asymmetric breast thicltenlng/
Asymmetric Fibrocysiic

 

 

 

 

 

Lump size:
Depth:
Hardness:
Mobility:
Shape:
Tamra: i Q- .jj Q I
Additional Findings With Lumps (chedt all that apply):
Skin DImpling/Retraclio Undocumented I
Skin Erythema {Undocumented "
a: sum: °I lamented
Nipple Retracli Undocumented ,
Nipple Scaling Undocumented
Pain/Tenderness Undocumented
Fibrocyslic Breast(s) Undocumented _
Nipple Discharge Undocumented
Domarsmdfl': _H ,., if
C] Nipple Disdlarge With No Lump
Spontaneous?
Color
Unilateral or bilateral?
Single or multiple ducts? I

 

 

 

Observational Findings With No Lump

 

CI Skin dlmpling/retraction

[:I Skin Erythema

III Skin Peau d'orange/Skln Thickening
[:1 Nipple retraction

E] Nipple scaling

 

 

 

 

[3?] Pa'" Dorcastpain

[:1 Chest wall pain
El Unspeciﬁed

 

 

 

C] Other, specify:

 

[:1 Lump(s)/mass(es)/Asymmetric breast thickening/
Asymmetric Fibrocystic

 

 

 

 

Lump size:
Depth:
Hardness:
Mobility:
Shape: .
Texture: I ____ h___ . I
Additional ﬁndings With Lumps (check all that apply):
Skin Dimpling/Retractio Undocumented
Skin Erythema Undocumented
$2 ﬁégnﬁgnge or Undocumented
Nipple Retraclion Undocumented ,
Nipple Scaling Undocumented
Painfl’endernees Undocumented
Fibrocystlc Breast(s) Undocumented
Nipple Discharge undocumented .,
[j Nipple Discharge With No Lump
Spontaneous?
Color
Unilateral or bilateral?
Single or multiple duds? I ’

 

 

E] Observational Findings With No Lump

 

CI Skin dimpling/retractlon

[:1 Skin Erythema

[:1 Skin Peau d'orange/Skin Thldtening
CI Nipple retraction

[:1 Nipple scaling

 

 

 

[2] Pain [:l Breastpain

[:l Chest wall pain
EM

 

 

 

Domenspecify: "

 

74

Quality of Written Description of CBE Documentation For Abnormal ﬁndings (Check All That Apply):

 

D Drawing of abnormal ﬁndings

 

 

 

C11 I . Nipple Change Undocumented I Breast Size/Sha Undocumented
I Star lUndocurnenmd I Skin Change {Undocumented

D Palpalion, specify: Fibrocystic Breast Undocumented ,_ Nodularity [Undocumented 7
Mass(es) Undocumented 7 Pain/tendemees Undocumented

 

C] Lymph node examination

 

 

Adenopathy/Axillary Nodes Undocumented Lymph Node Enlarged?

 

 

Dower, Specify:

 

 

 

“M 86’ Tl?” Go To Test ReSLiIt uh
I Followup F911} , I ..

 

75

Form III-Test Result Entry

Study to! . 1’ Date oruleVIsit: Lulu/L]

12. Mammogram Documentation:

 

 

 

 

 

1. Ordered/Recommended/Encouraged . ’ Date:
2.MammogramPerformed _ H 7‘ _ Date:
3. Results Obtained Stamped/Documented? :7 . Date:
4. Results Reviewed By FPCP Signed/Documented? j , . Date:
13a. Mammogram Findings: Final Impressions Which Breast? I
If you don't know which breast, please record Information in "Left Breast“ category.
Lel't Breast: Right Breast:
El Normal/No Finding mailed/Categoryl CI Normal] No Finding Identified/Camry!

[:l Normal/Benign-appearlng abnormality/Categor C] Normal/Benign-appearing abnormality/Catego
DProhaoiy benign/possibly malignant, inderterminate DProBaBly benign/possibly malignant, indeterminate

ICBIEQOFY 111 /Category III
DSusplclous for malignancy/Category IV [:ISuspicious for malignancy/Category IV
[:1 Malignant until proven otherwise/Categow V UMalignant until proven otherwise/Category V
DOther: Specify: , " L—JOther: Specify ‘”

 

 

 

13b. Mammogram Findings: Description Which Breast? A I“
If you don't know which breast, please record infonnatlon in 'Lel't Breast“ category.

 

 

 

 

 

Lel't Breast: Right Breast:
DAsymmetncBreastzmoreinwhichbreast " _" M
[:1 Bilateral Implants C] Bilateral Implants
E] Radioiucent Breasts E] Radioiucent Breasts
DDereeBreasis/DenseNoduiarBreasls DDenseBreasls/DenseNodularBreasts
CI Rounded density(les), most likely cyst or ﬂbroaden C] Rounded densities, most likely cyst or ﬁbroadeno
Cl Irregular Density(les) [:1 Irregular Density(les
[:l Benign Appearing Calciﬁcations CI Benign Appearing Caldlications
E] Suspicious Caldfication [:1 Susplcious Calcification
E] Calciﬁed Flbroadenomas [:1 Caldﬂed Fibroadenomas
E] Axillary Lymph Nodes [:1 Aldllary Lymph Nodes
13c. Mammogram Findings: Location For Category II and Up Whidl Breast? I

If you don't know which breast, please record information In "Left Breast“ category.
IF AREA NOT SPECIFIED, check SCATTER/THROUGHOUT Breast category
Left Breast Location: Right Breast Locaﬁon:

76

 

 

 

 

 

D Upper Outer Quadrant E] Lower Outer Quadrant [:1 Upper Outer Quadrant D Lower Outer Quadrant
D Upper Inner Quadrant E] Lower Inner Quadrant [:1 Upper Inner Quadrant C] Lower Inner Quadrant
D Lateral Breast E] Lateral Breast
[:1 Medial Breast E] Medial Breast
C] Areoiar/Nipple Area [:1 Areolar/Nipple Area
C] Deep Against Chest Wall E] Deep Against Chest Wall
E] Scattered/Throughout Breast C) Scattered/Throughout Breast
[:1 Other. specify: ‘ , D Other. specify: 7
14. Patient Notion of the Mammogram Findings? I .__,- _-__-.______. “i Date of Notiﬁcation:

ls--.- I -A-_,- ._... ., -.

15.(.yst-Fine Needle Aspiration (FNA)

Doneby: Datedone:

' i

 

V [:1 Mass resolved/ﬂuid not bloody [:I Fluid bloody

CIResidualMass
Clothe, specify:

 

DSentFiuldtoCytology

Rams Obtained
Results Reviewed By FPCP
Cytology Results:

Stamped/Documenmd?
Signed/Documented?

Date:
III Date:

 

C] Il'isufﬂcient/Hypocellular/Apoaine Cells
E] Atypical cells

 

[:1 Malignant

 

 

Domer.SDedfy:

 

C] Suspicious for malignancy [3 Benign/Fibrocystlc/Apocnne Cells

 

 

16. PatientNotil'iedoﬂiIeFNAFlndlngsFromCytology?

Date of Notiﬁcation:

17. Solid Mass-Fine Needle Aspiration Biopsy (FNAB)

 

DSpecimenSubmittedForAnalysis

Results Obtained Stamped/Documenmd?
Results Reviewed By FPCP Signed/Documented?
Pathology Results:

Date:
Date:

 

D Insuﬂ'ldent/Hypocellular

 

E] Benign/Fibrocystic
El Suspicious for malignancy [:1 Malignant

El Atypical cells

 

 

C] Other, specify:

 

18. PatientNotifiedofd'ieFNAB Findingsme Paleepolt?

 

Date of Notiﬁcation:

19. Ultrasound Findings:

Orderedby: Datedone:

Results Obtained

77

Stamped/Documented?

Date:

Result Reviewed By FPCP Signed/Documented? Date:

 

 

 

 

E] Negative ﬁnding [:1 Simple cyst(s) C] Solid mass(es) or complex cyst(s)
i:i Other.spe0fv' : ‘ ’ ‘ ‘ ‘ ‘
20. Patient untitled or the Ultrasound Findings? ’ ' ’ Date of Notiﬁcation:

—
21. Image-Guided Biopsy/Open Biopsy Result: Date done: i l i -

 

 

 

Result Received Stamped/Documented? ‘_ _ . Date:
Result Reviewed By FPCP Signed/Documentd? Date:
Open Biopsy Findings(check all that apply):

E] Benign/No Evidence of Malignancy [j Ductal Cardnoma in situ

E] Benign/Fibrocystic Changes E] Lobular Carcinoma in situ

E] Benign/Fat Necrosis El Atypical Hyperplasia

[:1 Benign/Lipoma C] Invasive Ductal Carcinoma

C] Benign/Fibroadenoma [:1 Invasive Lobular Carcinoma

[:I Other, specify: W ‘

 

 

 

 

 

 

 

uaaéaiatshvlngnigrlnmnw ﬂGnSI'Féi .. 'a..'i.
. Form ..E0||9wun .g

78

 

Form IV-Follow-up Entry

 

 

 

W0: L... ..-.. .. .1 Date ofVislt: Mini
23. Recommended Foliow-Up(s) (Check All mat Apply)
[3 Undocumented

Follow-up for Normal CBE and Mammogram (or One of Them Undocumented):

 

El Routine Screening [:1 12 Month CBE E] 12 Month Mammogram
[:1 Following Acs Guidelines El Following Other Guidelines specify:
Recommended by: I 7 ; Comment:

 

Follow-up for Specific Abnormalities: Follow-up To Any Abnormalities:

79

 

Bread Mass/Asymetry Initial Approach:

 

[:1 CBE at better phase cycle (3-10 days)
E] Fine Needle Aspiration for cyst

 

 

 

If Known Breast Cyst:

 

C] Send Fluid m cytology [:1 Reaspiration

EulowmanyhnonthCBE

 

C] Call if Problem Worsens

U Routine Screening

Recom. by:

Immediate Mammogram Workup:

 

 

 

If Known Solid Mass:

 

D Fine Needle Aspiration Biopsy

C] Specimen Submitted for Analysis

E] Repeat aspiration

1:] Clinical Followup Every 3 Months for 1 Year

 

 

[:1 Regular Mammociram

D Extra Mammogram Views
[:1 Cone or Spot Compression
C] Magniﬁcation Views

Recom. by:

 

 

 

For Nipple Discharge:

Interval Followup:

 

I I (How many) month mammogra

 

E] Endocrine work—up

 

 

l I (How many) month CBE

 

For Skin] Nipple Changes on Oburvatlon:

 

Recom. by: .

 

 

 

[:l 2 weeks antibiotics E] Skin Biopsy
C] Zweelttopicalhydrocort'sone

 

 

 

For Breast pain:

 

[:1 Eliminate Caffeine

[:1 Adjust Estrogen Dose

[:1 Local Anesthetic Injection

[:1 primrose Oil, How Many Months? E

[:l Reassurance and CBE within 3-6 months if pain persist
[j Supportive Brassiere

El Over-the—counter Analgesics

[:1 Danazol, Bromocriptine

 

 

 

For Occult Mammographic Abnomality:

 

[:1 Radioioglc Biopsy/Image-Guided Biopsy
Recommended by: I V 7

 

 

 

 

[:1 Ultrasound
Recom. by: .

Cl Surgical Referral
Recom. by:

[:1 Undocumented

Other Recommendations Or Comment
Consuming Abnomalltvﬂu):

 

 

 

 

.._, __ ,

General Comments About This Visit:

 

 

 

‘

8O

 

Assessment] Recommended Follow-up From Surgeon's Letter

 

 

1. Letter Written Date: I

2. Letter Received Stmped/Documented? Date:

3. Letter Reviewed by FPCP Signed/Dourmented? Date:
Assessment Followup

 

D Referral Diagnosis Not Conﬁrmed
Cl Referral Diagnosis Conﬁrmed
E] Additional] New ﬁndings

C] Further Test Recommended/Done By Surgeon, dieck all
that apply

[:1 Immediate Mammogra

[:1 Interval Mammogram, how long 3
Cl Interval CBE, how my:

D Ultrasound
[j FNA
Ci FNAB
C] Radiological/Image Guided Biopsy
El Open Biopsy
Evidence of Malignancy?
Ci Previous Abnormality Resolved
C] Current Abnormality Resolved
C] Other Comment From Surgeon's Lette

 

 

[:1 No Further Workup Required

C] Followup In Primary Care Office

[:1 Followup In Surgeon's Office

 

 

 

unus-

3' Go Back to

Form II

  
  

   

 

81

 

 

 

 

i

|

no Back to

 

Form III

 
   
   

Appendix 2

Kappa Calculation for Quality Control

TO perform the quality control we chose the relevant ﬁelds in the database for
which a kappa value could be calculated. The Kappa value is the ratio of the agreement
actually Observed minus the agreement expected by chance, divided by 1 (which
corresponds to perfect agreement) minus the agreement expected by chance:

K: (PA - Pc)/(1- Pc)

Kappa statistics were derived using the SAS program. The simple kappa
coefﬁcient measures the agreement between the abstractors beyond what could be
expected by chance.

Displayed below are three examples of the types Of Kappa calculations performed
on the data. These examples display the data collected, the SAS code used, and the
output produced by SAS.

Examples of Kappa calculation:

1. For ﬁelds with numerical value entries:
The following table is the data entered by both the abstractor and quality
control person for the question “Total numbers of visits within 15 months,
including the most recent visit” (question #3 on Front End Form). In this case
these numerical values were compared. In the table you will notice the
discrepancy between the abstractor and quality control for patient number 4.

 

 

 

 

 

 

 

 

 

Abstractor Quality Control
Patient 1 6 6
Patient 2 2 2
Patient 3 2 2
Patient 4 5 6
Patient 5 3 3
Patient 6 4 4
Patient 7 6 6
Patient 8 9 9

 

 

 

 

 

After this table is made, the data is input into SAS for Kappa calculation. The
Kappa result are the followings:

Kappa Statistics
Statistic Value ASE 95% Confidence Bounds

Silplo Kappa 0.8431 0.1430 0.5628 1.1234

Seaple Size = 8

83

2. Field labeled 0 or 1:
For ﬁelds with only 0 or 1 value, i.e. unchecked versus checked boxes
respectively, in the ACCESS Database, a different method of Kappa
calculation was used. An example of a scenario where this occurs is on form
II-Visit Entry. In this section the abstractors is asked to record CBE
documentation. One portion of the section is to indicate if the lymph node
examination is documented. The following table was made comparing the
abstractor versus quality control observations of whether during the CBE the
doctor documented a lymph node examination. In this example “1” signify
lymph node examination was documented and “0” means they it was not.

 

 

 

 

 

 

 

 

 

 

 

Abstractor Quality Control
Visit 1 O 1
Visit 2 0 0
Visit 3 O 0
Visit 4 O 0
Visit 5 O 0
Visit 6 0 0
Visit 7 1 1
Visit 8 1 1
Visit 9 O O

 

 

 

 

After this table is made, the data is transferred into SAS for Kappa
calculation. The Kappa results are the followings:

Sinple Kappa Coefficient

Sample Size = 9

3. Situations where Kappa is calculated to be 0%:
There are some ﬁelds with Kappa value equaling 0%. For these situations included
in parenthesis was the percent agreement. It has been documented and determined
by our study group that in some situations the Kappa statistics is not the best way to
represent the data and that in those situations the percent agreement is more
appropriate.

An example is included for bilateral mammogram ﬁndings. For a bilateral
mammogram, the abstractor is required to record mammogram ﬁndings for both
breasts. However, sometimes the abstractors would forget to record the bilateral
mammograms ﬁndings for one of the breasts.

The following table is the summary of bilateral mammogram documentation
results for several patients comparing quality control to the abstractor. In this case
“1” signiﬁes mammogram documentation and “0” signiﬁes no mammogram
documentation. In this scenario the abstractor missed recording the mammogram
documentation compared to the quality control for patient 4.

84

 

 

 

 

 

 

Quality Control Abstractor
Patient 1 1 1
Patient 2 1 1
Patient 3 1 1
Patient 4 l O

 

 

 

 

The Kappa results are the followings:

Simple Kappa Coefficient

Sample Size = 4

On the other hand, the percent agreement is calculated to be:
(4 — 1)/ 4 = 75%

85

“ma—mum “as gooﬂ H 2*: uuwoz

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

86

 

 

 

 

 

«$50 «$cw a «$ow Na

«$55 a a a _o

«$Vw * a * mm

* «$V5 «$ow a ~w

«$m5 * a «$mw ~5

«$5w a «$$w * No

«$mm «$ww * «$vw «c

a «$05 a «$mw Nm

* a a a «m

a * «$ow «$vw NV

«$65 «$V5 * «$oe «V

«$V5 «$mw * * NM

«$ww «$Vw * a —m

«$vw «$am a «$ac NN

«$wm «$N5 a «$Vc —N

* «$cm a a N~

«$ww «$55 a a -

arsesooam mg .58: a9 ES
Baa—0M 2 55:5 33> 3&0 “500M

26 seem ﬁsh .8 snaaz asap .8: 23 $8 £33m B §§B<

 

Ante...— aeuaﬁaeuﬂ 3.35va mar—eh 59:— 3—33— .255— : 03:.

 

28:58? .883 n A v

imam 89% $2: u ....... u3oz

 

$65

$05

 

a.
.33 o

*
$23 o

a\°
ex
<1-

o$~o

 

*

 

****

 

$55

goo

 

 

*
£35

 

 

******

.w
*3

 

$8

********

 

***l‘****

o\°
W
[\

***********

a.
$3

 

a.
fomc

i3 o

*

¢\°
o
00

$33

$3

 

a.

*

*

*****

 

*

.w

a.

.xﬁc

 

a.

 

*
:33 o

a.
$38 o

*
e\o5e

*

 

$w5

*****

*****O***********

*****************

*****

 

 

a.

2 5C

a 3::
38852

 

.w
A _ 5V

m 33
35:53..

A C 5v Exam
252 £993

 

25V

2283mm

 

39

553%:

 

 

nonﬁcoﬁaooa mmo

 

 

 

69
“08:00
mo 0&5

 

acuowbmn<

 

 

Ass... Pam Sit 55$ ES... 838m 2:5— a and.

87

EoEoouwm E083 u A V
H—zmﬂm MQQMVH goo" H 2*:

bowoumo

.5

H802

 

Nd

 

a

 

mm

 

«w

 

_5

 

No

 

G

 

mm

 

«m

 

gab

NV

 

$88

3

 

Nm

 

«m

 

NN

 

«N

 

mg

 

*****************

*****************

*****************

*****************

*****************

******O**********

*****************

*****************

*****************

*****************

*****************

*******O*********

:

 

dud
_> SD

 

co.“
> 30

 

to..—
>H EU

 

dog
a EU

 

«pa
a EU

 

and
a 30

 

23¢
H> EU

 

55
> so

 

25
a so

 

£5
B so

 

3mg
: so

 

:32
H .5

 

2%

 

 

$535k Eu SSS»:

 

 

is... him :33. .8: 5-8.8..— EEm £3.3— 332 a 2:5

88

“588%“ “coupon n A V

zaom Sag :52 n a... Munoz

 

Na

 

a

 

Nw

 

Aw

 

A5

 

No

 

_o

 

mm

 

*********

_m

 

**********

NV

 

on;

$°
\o
\o

3»

 

*

mm

 

£3

$43

gmw

Am

 

NN

 

***************

***************

AN

 

o
°\
00
[\

Ax.m5

NA

 

i
$38 o

****O************

{*ﬁ'l'ii'l-i'i'l-‘I'ilrl'i‘l-‘I

*****************

*******I*********

******w*****

****O************

c$5w

o$mw

:

 

 

_mbomom
seam

 

 

mmU
BEBE

 

.52

 

U)
.«E
>

«ham

 

0:552
338255

 

DEF—NE
5:2: 2

 

mmo
588 S

 

waﬁooaom
unused

 

3:88
-5825

 

e
2:.

 

 

AEeh mate—EB 3.83m SE..— m==8m «9:0— 3 03.5.

89

   
    

111111th

739

\I‘I‘ﬂl‘lﬂiﬂiﬂiﬂlﬂﬁlﬂ