"45818

(93,0 lull“!WillHIHHUIUIUIHIHillUlHHlllHlUUll

301835 4005

This is to certify that the

thesis entitled

Post-Discharge Seroma Formation Following Breast
Cancer Surgery: Implications for the Advanced
Practice Nurse

presented by

Mary J. Bloomfield

has been accepted towards fulﬁllment
of the requirements for

Master of Science degree in Nursing

 

 

2:] Z : 7
41— I’d Z" 5' /’ ﬂy
Major fessor

Date July 27, 1999

 

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

 

   

POST-DISCHARGE SEROMA FORMATION FOLLOWING BREAST CANCER

SURGERY: IMPLICATIONS FOR THE ADVANCED PRACTICE NURSE

By

Mary J. Bloomﬁeld

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
MASTER OF SCIENCE IN NURSING
College of Nursing

1 999

ABSTRACT

POST-DISCHARGE SEROMA FORMATION FOLLOWING BREAST CANCER
SURGERY: IMPLICATIONS FOR THE ADVANCED PRACTICE NURSE

By
Mary J. Bloomﬁeld

Medical management of breast cancer patients in the United States has changed
dramatically in recent years, largely in response to rising health care costs. A variety of
surgical procedures are available, including the traditional modiﬁed radical mastectomy
and new, breast-conserving techniques. Hospital stays of 24 hours or less have become
the norm, and patients are expected to perform more self-care actions during their
recovery at home. Among these actions is recognition of surgical complications, the most
common being seroma formation, a potentially infectious accumulation of ﬂuid in the
surgical site. The etiology of seroma is unknown, and prevention is problematic.

Results of this secondary analysis of the federal grant, A Subacute Care Nursing
Intervention for Short-Stay Breast Cancer Surgery (DAMD17-96-1-6325), Gwen Wyatt,
RN, PhD, Principle Investigator, suggest that women undergoing traditional surgery were
at ﬁve times greater risk for seroma formation compared to women receiving breast-
conserving surgery, regardless of the type of post-discharge care they received, or
personal characteristics such as body mass index or age. Implications for the Advanced
Practice Nurse (APN) on the interdisciplinary team include educating patients about
seroma formation, facilitating self-care, and collaborating with tertiary providers in order

to prevent serious sequella.

Copyright by
Mary Joanne Bloomﬁeld
1 999

This study is dedicated in memory of my mother, Marion Bloomﬁeld.

iv

ACKNOWLEDGMENTS

This author gratefully acknowledges Gwen Wyatt, RN, PhD (Chair), Brigid
Warren, RN, MSN, and Gail Casper, RN, PhD for their encouragement, support and
constructive assistance throughout the unfolding of this study.

This author also sincerely appreciates the support from my husband, family and

colleagues who offered encouragement during this process.

TABLE OF CONTENTS

LIST OF TABLES ....................................................... v.
LIST OF FIGURES ...................................................... vi.
ACKNOWLEDGMENTS ............................................... vii.
INTRODUCTION
Statement of the Problem ............................................ 1
REVIEW OF THE LITERATURE
Risk of Seroma formation following breast cancer surgery .................. 6
Variables associated with seroma formation identiﬁed in the literature ........ 10
Summary and Critique of the Literature Review ......................... 14
Discussion of the Rationale for Proposed Study .......................... 15
CONCEPTUAL FRAMEWORK .......................................... 17
CONCEPTUAL DEFINITIONS
Seroma ......................................................... 20
Type of Post-operative Care .......................................... 21
Type of Surgery ................................................... 21
Body Mass Index (BMI) ............................................ 21
Age ............................................................ 21
METHODS
Sample .......................................................... 21
Parent Study Design ............................................... 22
Hypothesis and Research Questions .................................... 23
Secondary Analysis Design .......................................... 24
Procedures for Data Collection ....................................... 26
Procedures for the Protection of Human Subjects ......................... 26
Operational Deﬁnition of the Variables
Seroma ................................................... 26
Type of Post-operative Care ................................... 27
Type of Surgery ............................................. 27
Body Mass Index (BMI) ....................................... 27
Age ....................................................... 27
DATA MANAGEMENT ................................................. 27

vi

RESULTS

Description of Sample Demographics ................................. 29
Answers to Research Questions ...................................... 30
DISCUSSION
Interpretation of the Findings ......................................... 36
Discussion of Results with the Conceptual Model ........................ 37
Implications for the Advanced Practice Nurse ............................ 38
Recommendations for Further Research ................................ 39
Summary ........................................................ 40
APPENDIX ............................................................ 41
REFERENCES ......................................................... 42

vii

LIST OF TABLES

TABLE 1: Frequency of Type of Surgery and Seroma Formation .................. 29
TABLE 2: Demographics ................................................. 30
TABLE 3: Main Effect Model .............................................. 31
TABLE 4: Testing for Confounding Variables ................................. 32

viii

LIST OF FIGURES

Figure 1: Dorothea Orem’s Conceptual Framework for Nursing (1995) ............. 17

Figure 2: Adapted from Dorothea Orem’s Conceptual Framework for Nursing ....... l9

ix

Introduction
Statement of the Problem

In response to rising health care costs, surgical management of breast cancer
patients has changed dramatically in recent years. Hospital stays of 24 hours or less have
become the norm, and patients are increasingly expected to engage in self-care actions
during their recovery period at home (Clark & Kent, 1997; Karnbouris, 1996). These
actions include monitoring for post-surgical complications.

The most frequent complication after breast cancer surgery is seroma formation.
Although not as serious a complication as frozen shoulder or lymphedema, the problem
of seroma formation is worthy of study by the Advanced Practice Nurse (APN) in the
current health care environment. A seroma is ﬂuid that accumulates in surgical dead
space which creates a suitable culture medium for retrograde skin bacteria in the presence
of a foreign body, such as a drain. Infection and sepsis can result as long as the drain
remains intact. If the drain is removed prematurely, a seroma may form, and must be
treated with one or more percutaneous aspirations, reinsertion of a drain, or open drainage
by lancing. The possible etiologies of seroma formation are unknown.

Data from the federal grant, A Subacute Care Nursing Intervention for Short-Stay
Breast Cancer Surgery (DAMD17-96-1-6325) (Wyatt, Given, & Given, 1995), were used
to analyze a collection of risk factors associated with seroma formation. Begun in 1996,
the parent study was a four-year federally funded nursing intervention program that
sought to determine the type of care women needed at home after short-stay surgery, and

taught women how to care for themselves post-discharge.

Shortened hospital stays mandate that patient care be performed by the patient
and/or family caregivers. The APN is uniquely qualiﬁed to facilitate patient self-care. As
part of an interdisciplinary health care team planning post-discharge care, the APN
provides continuity of care throughout the disease process by collaborating with other
health care providers.

As a primary care provider, the APN screens patients for breast cancer is the ﬁrst
health care provider to educate patients about self care needs. In the event that breast
cancer is found, the APN in primary care provides preliminary teaching about the disease
process and treatment options, and refers patients to tertiary care providers.

As part of the interdisciplinary health care team planning post-discharge care, the
APN is concerned with post-surgical complications, the most common being seroma
formation. Seromas usually occur one to four weeks aﬁer the patient has been discharged
ﬁom the hospital (Clark & Kent, 1992; Holcombe, West, Mansel, & Horgan, 1995;
Burak, Goodman, Young, & F arrar, 1997). Unless the seroma is recognized and treated
promptly, it can become infected, interfere with wound healing, delay adjuvant cancer
therapy, and increase patient discomfort and health care costs. For the APN, knowledge
about possible risk factors for seroma will improve the quality of patient education,
thereby facilitating optimal self-care throughout the disease process.

The nursing protocol administered to women in the intervention arm of the parent
study took place during the ﬁrst two weeks after surgery. Nurses instructed patients to
engage only in light activities of daily living while the drain was intact, and to measure

and record drainage output daily. Approximately one week later, patients were instructed

to begin range of motion exercises on a daily basis, gradually increasing activity until pre-
operative range of motion was recovered (American Cancer Society, 1996). The main
objective was to restore range of motion through patient self-care during the ﬁrst two
weeks after breast cancer surgery. Participating surgeons performed both traditional and
breast-conserving surgery. All patients were discharged from the hospital within 48 hours
with a drain that remained intact for at least three days after surgery.

The shift to short-stay breast cancer surgery moves post-surgical care from the
tertiary health care setting into to the patient’s home. The interdisciplinary health care
team works together to avoid treatment delays between primary, secondary, and tertiary
care. The APN is uniquely qualiﬁed to collaborate with physicians and nurses in order to
identify and treat seroma formation, as well as to provide the patient education and
referral process necessary to prevent serious sequella (Dontje, Sparks, & Given, 1996).

According to van Wersch, Bonnema, Prinsen, Pruyn, Wiggers, and van Geel
(1997), a majority of breast cancer patients believed that they were poorly informed about
how to receive help and guidance following discharge from the hospital, and wanted
greater interaction between health care professionals. Continuity of care for breast cancer
patients can be achieved when the APN, in collaboration with other health care providers,
helps the patient meet her health care needs throughout the disease process.

The purpose of the secondary analysis was to determine whether a combination of
possible risk factors identiﬁed in the literature predicted seroma formation. Individual
variables were introduced simultaneously. Research questions for this secondary analysis

include the following:

1) Does type of post-discharge care predict seroma formation?

2) Does type of surgery predict seroma formation?

3) Does patient Body Mass Index (BMI) predict seroma formation?

4) Does patient age predict seroma formation?

Seroma formation was deﬁned as an accumulation of serous and/or lymphatic
ﬂuid in the surgical dead space resulting from the surgical removal of breast tissue and/or
axillary lymph nodes that required follow-up treatment in the form of aspiration, lancing,
or drain reinsertion. Type of post-discharge care was deﬁned as (1) the parent study’s
post-discharge nursing intervention or (2) post-discharge care provided by a nursing
agency, family members, friends, or the patient herself.

Participants in the intervention arm of the parent study received a two-week
nursing protocol consisting of an average of three home visits and ﬁve phone calls
designed to facilitate patient self-care and compliance with the post-discharge care plan.
Some participants in the control arm received daily visits from a visiting nurse, usually
over a one week period. Unlike the parent study’s nursing intervention, visiting nurses
focus a series of tasks, rather than enabling the patient to perform self-care. Women in the
control arm received post-discharge care from a family member, friend or the patient
herself. The hypothesis is that women receiving post-discharge care from a source other
than the parent study’s intervention nurse will be at greater risk for seroma formation.

The trend toward breast-conserving surgical methods is fairly recent, and its effect
on seroma formation has not been widely studied. Seroma rates among women who had

breast-conserving surgery, deﬁned as lumpectomy with axillary node dissection, were

compared with the rate of seroma experienced by women who underwent traditional
modiﬁed radical mastectomy or wide excision with axillary lymph node dissection. The
hypothesis is that women who have traditional surgery are at greater risk for seroma
formation, compared to women who undergo the less invasive breast-conserving surgery.

Two demographic variables, patient Body Mass Index (BMI) and age, also were
considered in the secondary analysis as possible risk factors for seroma formation. BMI
accounts for differences in body composition by deﬁning the level of adiposity according
to the relationship of weight to height. The hypothesis is that patients with high levels of
adiposity (obesity) are at greater risk to develop a postoperative seroma than women with
ideal or low BMI scores. Patient age was measured in years, with the hypothesis being
that older women are at greater risk for seroma formation than younger women.

Review of the Literature

Risk of Smma Formation Following Breast Cancer Sggeg

Traditionally, deﬁnitive breast cancer surgery involved a two-stage diagnostic
biopsy and therapeutic surgical procedure done on an inpatient basis, with hospital stays
ranging from 23 hours to several days. However, recent surgical literature explores the
effect of breast cancer surgery done on an outpatient basis, with the conclusion that it
offers safe and cost-effective cancer management. The year 1991 was the ﬁrst year that
the total number of outpatient breast cancer surgical procedures surpassed the number of
inpatient procedures in the United States (Goodman & Mendez, 1993).

In 1993, Goodman and Mendez conducted a retrospective review of 221 patients

ranging in age from 34-87 years who, on an outpatient basis, underwent total mastectomy,

modiﬁed radical mastectomy, partial mastectomy with and without axillary node
dissection, and axillary node dissection only. Drains were removed between 5 and 7 days
postoperatively. The incidence of seromas requiring one or more aspirations was less than
15%, comparable with the incidence in patients operated on as inpatients.

Similarly, McManus, Topp, and Hopkins (1994) conducted a retrospective
analysis of 118 patients ranging in age ﬁ'om 29 to 87 years, who, on an outpatient basis,
underwent modiﬁed radical mastectomy, lumpectomy and axillary node dissection, and
axillary node dissection alone. Patients who had undergone axillary dissection had their
drains removed in 48 hours, while patients who had undergone mastectomy had their
drains removed in 5-7 days. The study revealed that 87% (n = 90) of patients undergoing
lumpectomy and axillary dissection did so in the outpatient setting, while 36% (n == 20) of
modiﬁed radical mastectomy patients were treated on an outpatient basis. Although
several seromas were managed by aspiration, only three patients required post-surgical
hospitalization. The investigators concluded that same-day discharge was safe.

Post-surgical seroma remains a quandary because its etiology is unknown.
Although seroma is generally deﬁned as the accumulation of excess ﬂuid in the surgical
dead space, deﬁnitions in the literature vary. For example, Kumar, La], and Misra (1995)
found a 28% incidence of seroma formation, deﬁned as an accumulation of serous ﬂuid
underneath the mastectomy ﬂap. Petrek, Peters, Cirrincione, and Thaler (1992), stratiﬁed
subjects according to axillary dissection or modiﬁed radical mastectomy, and deﬁned
seroma as the accumulation of lymphatic ﬂuid in the axillary space and accumulation of

serum under the mastectomy ﬂap. Holcombe, West, Mansel and Horgan (1995) found a

50-60% incidence of seroma requiring aspiration post axillary node dissection, but did
not stipulate whether ﬂuid was lymphatic or serous. In their study involving both
traditional and breast-conserving surgical techniques, Burak, Goodman, Young, and
Farrar (1997) found a 40% seroma rate overall, and concluded that seroma formation
occurs as a consequence of both lymphatic disruption and oozing of capillary beds.
Sommers, Jablon, Kaplan, Sandler, and Rosenblatt (1991) deﬁned seroma as a
palpable collection of ﬂuid in the axilla, and concluded that use of a closed suction
drainage system after lumpectomy and axillary node dissection was advantageous in
decreasing the degree of seroma formation. They randomized 227 patients scheduled for
axillary node dissection into a drain group (108 cases) and a no drain group (119 cases).
All drains were removed on the ﬁrst postoperative day. Overall, 81.5% of patients
developed postoperative seromas. In the drain group, 73.2% of patients required
aspirations compared with 89.1% of patients in the no drain group. The mean number of
aspirations in the drain group was 2.2, with a range of 0 to 11, and 3.3 in the no drain
group, with a range of 0 to 11. These differences were found to be statistically signiﬁcant
(p = 0.002) using a two-sample t-test. Mean volume aspirated in the drain group was less
than the no drain group (146.3 ml versus 266.1 ml, p = 0.003). Length of morbidity was
11.5 days in the drain group, statistically smaller than the no drain group (p = 0.0002).
Similarly, a study of 86 surgical patients by Holcombe, et a1. (1995), found that
frequency of postoperative seroma formation was signiﬁcantly reduced for both
mastectomy and breast conservation patients when drains were maintained for a longer

duration of time [7 cases (18%) versus 16 cases (35%), P < 0.001 ; X2]. Dietrick-Gallagher

and Hyzinski (1989) also found some evidence that seroma frequency is negatively
correlated to the amount of time drains are maintained. Their study compared randomly
assigned groups of patients who were discharged from the hospital before drain removal
(n = 15), and after drain removal (n = 22). While a t-test revealed no statistically
signiﬁcant differences in complication rates between the groups, the rate of seroma
development was higher in the hospitalized partial mastectomy group. The investigators
speculate that this ﬁnding may be related to the fact that hospitalized patients had
drainage systems intact for an average of only 3.72 days, compared with 6.25 days for the
early discharge group.

In contrast, Yii, Murphy, and Orr (1995), conducted a trial with 100 consecutive
patients undergoing modiﬁed radical mastectomy or breast conserving surgery, and
concluded that seroma formation is not signiﬁcantly associated with the amount of time
the drain remained intact for either type of surgery. Seroma was deﬁned as ﬂuid
accumulation in the surgical site requiring aspiration. Factors considered in this study
included drainage volume within a 48 hour period, total drainage volume, and amount of
time closed suction drains remained intact between two randomly assigned groups of
patients. One group had drains removed four days after surgery, and the other group had
drains removed 48 hours after surgery. Seromas developed in three cases (6%) in the
group with drains in place for four days, compared to 5 cases (10%) among the group
with drains in place for 48 hours, a ﬁnding that does not represent a signiﬁcant difference
between the two groups. The investigators concluded that drains may be removed on the

second postoperative day without increased risk of seroma formation.

Similarly, a study of 98 modiﬁed radical mastectomy patients by Parikh, et a1.
(1992), suggests that seroma formation, deﬁned as an accumulation of ﬂuid in the
surgical site that required aspiration, is not associated with the amount of time drains are
maintained after surgery. Patients were randomized into one of two groups: one group
had early drain removal (on the third postoperative day), while the second group had late
drain removal (on the sixth postoperative day). Neither mean volumes of seromas
aspirated (p = 0.799) nor in the number of aspirations (p = 0.740) were statistically
different between groups, and the investigators concluded that drains may be removed by
the third postoperative day. The one factor shown to inﬂuence seroma volumes was body
weight greater than 65 kilograms. Height was not included as a factor.

Typ_e of Post-Discharge Care

The literature is limited regarding the effect of post-discharge nursing protocol on
seroma formation among short-stay breast cancer patients. Over the past decade, health
care delivery in the United States has changed dramatically, and the results of those
changes can now be analyzed. Outcome measures such as seroma formation provide a
yardstick by which to evaluate the effects of economic and organizational change in the
health care system on the breast cancer patient. Thus, the impetus for this study.

In 1989 at the Hospital of the University of Pennsylvania in Philadelphia,
Dietrick-Gallagher and Hyzinski implemented an early discharge education protocol for
15 mastectomy patients that included pre-discharge nursing instruction about incision and
drain care, a discharge instruction sheet, and a postoperative office visit. Rate of seroma

formation among the early discharge group was then compared with the rate of seroma

among 22 randomly selected mastectomy patients who remained hospitalized until the
drain was removed. The groups were similar in age, had the same number of lymph nodes
removed, and had similar numbers of positive lymph nodes. A t-test was used to analyze
the data, and no signiﬁcant difference in seroma formation was found between groups.
Among the early discharge group, the rate of seroma formation was 29%, compared to
30% among hospitalized patients.

Holcombe, et al. (1995) also compared hospital nursing care with home nursing
care in a pilot study at the University of Wales College of Medicine, Cardiff, United
Kingdom. One hundred and one women who had a modiﬁed radical mastectomy or wide
excision and axillary lymph node dissection as part of their breast cancer treatment were
studied. The mean age was 59 years within a range of 29-76 years. Thirty-nine patients
opted for early discharge, which was, on average, ﬁve days shorter than the control group.
The early discharge plan was explained during preoperative counseling, and included an
instruction and an information sheet about care of the drain. Investigators found that
patients in the intervention group had lower rates of seroma formation when compared to
patients who received more days of hospital care (18% vs 34%, p = 0.001; X‘). Both
studies comment on the fact that patients who recovered at home tended to keep drains
intact for a longer time period than women who recovered in the hospital.

In a study of 208 breast cancer patients from his practice at Henry Ford Hospital
in Detroit, Michigan, Karnbouris (1995) found that approximately 45% of participants
were treated with modiﬁed radical mastectomy, 35% with breast-conserving surgery, and

the remaining with a variety of procedures tailored to speciﬁc circumstances. All

10

participants underwent complete axillary node dissection on an outpatient basis (hospital
stays of 24 hours or less) and were permitted unrestricted use of arm and shoulder aﬁer
surgery. Thirty-two percent of the participants developed a seroma; however, the study
design did not include a control group for comparison.

All three of the studies above emphasized the cost savings when length of hospital
stays are reduced. The majority of patients in each study underwent traditional modiﬁed
radical mastectomy. Seroma incidence varied between 18% and 32% for patients who
opted for short-stay surgery, while patients who had a longer hospital stays were found to
have seroma rates between 30% and 34%.

Type of Surgeg

The literature suggests that seroma formation is common after all types of breast
cancer surgery (Yii, Murphy & Orr, 1995; Holcombe et al., 1995; Kumar, et al., 1995);
however, the risk of seroma formation among patients who opt for breast conserving
surgery as opposed to modiﬁed radical mastectomy or wide excision is unclear. In the late
1980's and early 1990's, several prospective, randomized trials demonstrated that survival
after lumpectomy, axillary dissection, and radiotherapy was equivalent to survival after
mastectomy (Morrow & Hassett, 1994; Fisher, Redmond, & Poisson, 1989; Veronesi,
Banﬁ, & Del Vecchio, 1986). As a consequence, breast conserving surgery has become
the therapy of choice, unless speciﬁcally contraindicated. However, few studies take into
account the impact of the surgical method on seroma formation.

In 1985, Tejler and Aspergren conducted a study involving 385 patients to

compare complications after modiﬁed radical mastectomy (324 cases or 82%) and breast

ll

conserving surgery, deﬁned as segmental resection with axillary dissection (71 cases or
18%). They found seromas in 36.5% of wounds after modiﬁed radical mastectomy, and in
26.8% after segmental resection. In addition, seroma formation was positively correlated
with increasing age (p 5 0.05).

More recently, Burak, Goodman, Young, and Farrar (1997), conducted a
prospective, randomized study of 101 patients who underwent modiﬁed radical
mastectomy or breast conserving surgery, deﬁned as lumpectomy with axillary node
dissection. The mean age was 54 (range 31-86 years), average weight was 68.9 kg (range
40-140 kg), and the average Quitelet Obesity Index was 26.5 (range 17-54). Using
stepwise logistic regression, ﬁndings indicated that breast-conserving surgery (p = 0.001),
increased patient age (p = 0.016), increased patient weight (p = 0.122), and high amounts
of initial 72-hour wound drainage (p = 0.001) were risk factors for seroma formation.
Patient BMI and Age

Many researchers have considered the impact of demographic variables, most
often patient weight and age, on seroma formation. In the literature, patient weight is
usually deﬁned as number of pounds or kilograms. As mentioned earlier, Parikh, et al.
(1992), found a positive correlation between patient weight and seroma formation.

Similarly, Kumar, et al., (1995) reported a positive correlation between seroma formation
and both weight and age in a prospective study involving a convenience sample of 64
female patients undergoing modiﬁed radical mastectomy. Patients who developed
postoperative seroma were designated as group A (n=18, or 28%) and patients who did

not develop seroma were designated as group B (n=46, or 72%). Group A patients were

12

signiﬁcantly older (p = 0.02) and heavier (p 5 0.05). However, the investigators note that
a signiﬁcantly greater number of group A patients were hypertensive, and that the age and
weight factors should be viewed as factors inﬂuencing the cardiovascular status of the
patients rather than being determinants of seroma formation.

As mentioned earlier, Burak, et al., (1997), suggest that obesity and increasing age
put women at risk for seroma formation. In 1985, Tejler and Aspegrem followed 313
modiﬁed radical mastectomy patients and found 116 seromas (37.1%). The occurrence of
seroma was positively correlated with increasing age. However, a retrospective analysis
of 351 modiﬁed radical mastectomy patients by Chilson, Chan, and Lonser (1992) found
no association between patient age and seroma.

Cri i of the Literature Review

Review of the literature reveals contradictory ﬁndings, which indicates that
seroma formation has a non-linear relationship with factors such as age, weight, and type
of surgery. For the most part, research was conducted by surgeons who, in response to the
rising cost of hospitalization, were trying to identify risk factors for seroma formation in
hopes of reducing the number of days patients remained in the hospital aﬁer surgery.
Statistical techniques that test the association between variables, such as t-tests and chi-
square, were most often used. Several studies were based on very small sample sizes,
which diminished the veracity of their conclusions and contributed to the contradictory
ﬁndings evident in the literature overall.

Surgical practice varies slightly, but there is general consensus that a closed-

suction drainage system maintained for two to seven days after surgery is standard

13

practice (Dietrick-Gallagher & Hyzinski, 1989; Holcombe, et al., 1995; Parikh et al.,
1992; Sommers et al., 1991; Yii, et al., 1995). Studies questioning the effect of treatment
variables, such as type of surgery and post-discharge care, are sparse, despite the
widespread adoption of breast-conserving surgery and shorter hospital stays.

The impact of nursing care on seroma formation is seldom addressed in the
literature, and is often limited to comparisons of hospital nursing care versus home
nursing care. Studies looking at the impact of demographic variables such as patient age
and weight on seroma formation also have been poorly deﬁned and have yielded
contradictory ﬁndings. Therefore, the APN will beneﬁt from further study of these
variables in order to identify possible risk factors for seroma formation, and to improve
patient self-care following breast cancer surgery.

Discussion Ed Rationale for Promsed Study

The secondary analysis of A Subacute Care Nursing Intervention for Short-Stay
Breast Cancer Surgery (DAMD17-96-1 -6325) utilized logistic regression to analyze the
impact of four independent variables - type of post-discharge care, type of surgery, patient
BMI, and patient age - as a collection of factors that, in and of themselves, may not be
signiﬁcant, but in relation to the other variables may be found to have signiﬁcant
predictive value in regard to seroma formation. Variables found to predict seroma
formation were then considered within the context of patient education and continuity of
care for patients with breast cancer who are recovering from surgery at home.

In response to recent changes in the surgical treatment of breast cancer patients,

the secondary analysis examined the ﬁequency of seroma formation among women who

14

had breast conserving surgery, deﬁned as lumpectomy with axillary node dissection,
compared to women who had a modiﬁed radical mastectomy or wide excision with
axillary node dissection, in order to indicate whether risk of seroma formation is
associated with the type of surgical procedure experienced by the patient.

The weight variable was considered in the more explicit terms of Body Mass
Index (BMI). This index accounts for differences in body composition by deﬁning the
level of adiposity according to the relationship of weight to height, and has the least
correlation with body height and the highest correlation with independent measures of
adiposity for adults, including the elderly. BMI is deﬁned according to the following
scoring method: (A) under 20 is associated with underweight and may be associated with
health problems for some individuals; (B) 20—25 is the "ideal" range associated with the
lowest risk of illness for most people; (C) 25-27 is associated with overweight and may
be associated with health problems for some people; (D) 29 or greater for women
indicates obesity (Wardlaw & Insel, 1993).

Patient age has long been suspect as a variable associated with seroma formation.
For the secondary analysis, age was measured in years as a continuous variable.

Short hospital stays following breast cancer surgery is a fairly recent phenomenon,
and its impact on recovery could be substantial. Patients recovering at home require much
more instruction about post-discharge care and the risk of potential surgical
complications. As post-surgical care moves out of the hospital and into the patient’s
home, the APN is uniquely qualiﬁed to collaborate with tertiary health care providers in

order to ensure continuity of care for the patient with a diagnosis of breast cancer.

15

Conceptual Framework
Dorothea Orem’s Theory of Self-Care

In today’s health care delivery system, women are usually discharged from the
hospital within 48 hours aﬁer breast cancer surgery, and are more reliant on self and
family for post-surgical care that was previously provided by hospital nurses (Clark &
Kent, 1997; Kambouris, 1996). This secondary analysis seeks information about possible
risk factors associated with seroma formation, and about how that information can be
employed by the APN on an interdisciplinary health care team planning post-discharge
care. Dorothea Orem’s self-care emphasis can be used by the APN to ensure minimal
provider intrusion with maximum patient responsiveness.

Dorothea Orem’s conceptual model of nursing focuses on individuals’ self-care
needs and their capabilities for meeting those needs. According to Orem, the goal of
nursing is to help people meet their needs at the therapeutic level and on a continual basis
(Orem, 1995). People are considered legitimately in need of nursing care when a self-care
deﬁcit is diagnosed by the nurse. Orem’s major theoretical concepts are: (1) conditioning
factors; (2) health deviation; (3) self-care requisite; (4) self-care; (5) therapeutic self-care
demand; (6) self-care deﬁcit; (7) nursing agency; and (8) self-care agency (see Figure 1).

This secondary analysis attempted to identify possible risk factors for seroma
formation, as a way for the APN to facilitate self-care. As a consequence, Orem’s ﬁrst
three concepts - conditioning factors, health deviation, and self-care requisites - are

considered relevant to this study (see Figure 2).

16

Conditionin Factors: Health State and E ect a Medical Care

\.

Health Deviation: Deviation rorn normal structure or auction

\.

Self-Care R uisites: De med b Health Deviation

 

 

 

Self—Care Therapeutic

Agency Self-Care Demand

 

Figure 1. Dorothea Orem's conceptual framework for nursing
(R = relationship;< = deﬁcit relationship, current or projected), Orem (1995).

Conditioning Factors
Orem deﬁnes conditioning factors as the patients health state, the effects of

medical care, and environmental conditions that predispose the patient to a particular

health deviation. In this secondary analysis, conditioning factors are the four risk factors

17

under study. Two factors, namely type of post discharge care and type of surgery, involve
the effects of medical care on seroma formation. The last two factors - patient BMI and
age - consider the effects of the patient’s health state on seroma formation (see Figure 2).
Health Deviation

A health deviation is an illness or other aberration from the normal structural and
functional integrity of the patient. The health deviation determines what kind of self-care
requirements must be met by the patient. In the secondary analysis, seroma formation is
the health deviation under consideration (see Figure2).
Self-Care Reguisites

Orem identiﬁes three types of self-care requisites: (1) universal self-care requisites
that involve activities such as adequate intake of air, water, and nutrition in the
maintenance of the human structure and its functioning; (2) developmental self-care
requisites that involve successful passage through various stages in the life cycle; and (3)
health-deviation self-care requisites that involve the actions that must be performed by
the patient in response to a particular medical treatment modality. In this secondary
analysis, the self-care requisite is deﬁned by the health deviation (seroma formation) as
prevention of serious sequella related to seroma formation (see Figure 2).
Conceptual Deﬁnition of Seroma

A seroma was deﬁned in this study as an accumulation of serous and/or lymphatic
ﬂuid in the surgical dead space that required treatment in the form of aspiration, lancing,

or reinsertion of a drain. Within the context of Orem’s theory, seroma formation is

18

Conditionin Factors: T e o Post-Dischar e Care T e 0 Su e BMI A e

\.

Health Deviation: Develo ment 0 a seroma

\.

Self-Care Re uisites: Prevent serious se uella related to seroma orrnation

 

 

 

      
   
     
 

   

Self-Care
Agency

Therapeutic
Self-Care Demand

 

   

  

PATIENT

Nursing
Agency

Figure 2. Adapted from Dorothea Orem's conceptual framework for nursing
(R = relationship; < = deﬁcit relationship, current or projected).

deﬁned as a health deviation that mandates particular self-care requisites, namely,
prevention of serious sequella related to seroma formation.
Conceptual Deﬁnition of Type of Post-Discharge Care

For the secondary analysis, post—discharge care was deﬁned as either the parent
study's nursing intervention or conventional care provided by a nursing agency, friends,

19

family, or the patient herself. According to Orem’s model, type of post—discharge care is
deﬁned as a conditioning factor that results from the effects of medical care.
Conceptual Deﬁnition of Tm pf Surgm

In the secondary analysis, type of surgery was limited to breast conservation,
deﬁned as lumpectomy with axillary node dissection, and traditional surgery, deﬁned as
modiﬁed radical mastectomy (removal of breast and axillary lymph nodes) and simple
mastectomy (removal of breast only). Under Orem’s model, type of surgery is categorized
as a conditioning factor related to the effect of medical care.

Cpnceppﬂ Deﬁnitipn of BMI and Age

BMI was deﬁned as a combination of the patient’s height and weight, as outlined
in Section 2. In Orem’s model, patient BMI was categorized as a conditioning factor
related to the patient’s health state. Age in years was calculated for adult women in the
study, and also was deﬁned as a conditioning factor related to the patient’s health state.

Methods
mm:

This study is a secondary analysis of 102 participants from the federal grant, A
Subacute Care Nursing Intervention for Short-Stay Breast Cancer Surgery (DAMD17-
96-1-6325), Gwen Wyatt, RN, PhD, Principle Investigator. The convenience sample was
drawn from women with known carcinoma of the breast residing and receiving surgical
care in the State of Michigan. Eligibility criteria for participants in the parent study were:
(1) age 21 or older, (2) conﬁrmed diagnosis of breast cancer, (3) surgical treatment that

included placement of a closed-suction drain, and (4) current residence within 40 miles of

20

the surgeon’s office. Exclusionary criteria were: (1) pregnancy, (2) active psychiatric
disorder, (3) primary language other than English, (4) postoperative hospital stay longer
than 48 hours, (5) residence in a nursing home or assisted living facility, (6) in situ breast
tumors, and (7) reconstructive surgery concurrent with removal of cancerous tissue.

All participants underwent lumpectomy with axillary lymph node dissection,
modiﬁed radical mastectomy with axillary lymph node dissection, wide excision with
axillary lymph node dissection, or simple mastectomy without axillary lymph node
dissection. Patients stayed in the hospital 48 hours or less after surgery and returned home
with a closed suction drain.

Parent Study Desig

The parent study, A Subacute Care Nursing Intervention for Short-Stay Breast
Cancer Surgery (DAMDI 7-96-1-6325), followed an experimental design and assigned
participants to one of two arms: (1) the intervention arm, in which participants received a
nursing protocol that emphasized self-care; and (2) the control arm, which included
patients receiving conventional care from a community nursing agency, care by a friend
or family member, or self-care. Participants were randomly assigned to either the
intervention arm or the control arm of the study. Women assigned to the intervention arm
received a self-care nursing protocol from the parent study for a two-week period
immediately after surgery. The control group received conventional post-discharge care in
the home for a one-week period from a community home nursing agency, or care was
provided by a friend, family member, or the patient herself. Random assignment was

carried out in groups of ten to ensure equal distribution of subjects at any point in time.

21

In the parent study, breast cancer surgery was deﬁned as: (1) lumpectomy with
axillary lymph node dissection; (2) modiﬁed radical mastectomy; (3) wide excision with
lymph node dissection; and (4) simple mastectomy without lymph node dissection. Study
recruiters obtained written consent from patients in Lansing, Charlotte, Pontiac, and Troy
Michigan. Participants were randomly assigned to an intervention group, who received
the study’s nursing care in the home for a two week period aﬁer surgery, or a control
group, who received nursing care in the home by various home health agencies, by family
members, or by the patient herself. Four weeks after surgery, participants were
interviewed about their recovery. The goal of the study was to determine the optimal
amount and type of home nursing care needed after short hospital stays.

The nursing protocol administered to women in the intervention arm of the parent
study focuses on self-care. Study nurses instructed patients to engage only in light
activities of daily living while the drain was intact, and to maintain their drains until total
24-hour drainage was under 20-30 ml for three consecutive days. After the drain was
removed, participants were taught range of motion exercises (American Cancer Society,
1996) and encouraged to perform them daily, gradually increasing the number of
repetitions until preoperative range of motion was restored. The protocol was designed to
facilitate patient self-care by providing instruction and information about incision and
drain care, the risk of postoperative complications, and the disease process.

S con a1 sis: Desi
Secondary analysis involves the use of data gathered in a previous study to test

new hypotheses or explore new relationships (Polit & Hungler, 1995). It is an efﬁcient

22

and economical design because it analyzes data that the parent study has already
collected, but did not examine in primary research questions. The use of available data
makes it possible to bypass time-consuming and costly steps in the research process.

Secondary analysis of A Subacute Care Nursing Intervention for Short—Stay
Breast Cancer Surgery (DAMD17-96-1-6325) examines possible risk factors associated
with seroma formation, and considers the implications for the APN on an inter-
disciplinary health care team planning post-discharge care. The dependent variable,
seroma, was linked to four antecedent independent variables: (1) type of post-discharge
care, (2) type of surgery, (3) patient Body Mass Index (BMI), and (4) patient age.

The secondary analysis asks whether type of post-discharge care predicts seroma
formation. The two arms of the parent study were used to deﬁne the independent variable.
The hypothesis is that patients receiving conventional care by a visiting nurse, family,
friends, or the patient herself, are at greater risk for seroma formation compared to
women participating in the parent study’s nursing intervention. The second hypothesis is
that women undergoing traditional breast cancer surgery are at greater risk for seroma
formation than those receiving breast-conserving surgery. The third hypothesis posits that
obesity, as deﬁned by the Body Mass Index, is a risk factor for seroma formation, and the
fourth hypothesis is that elder patients are at greater risk for seroma formation.

The major disadvantage of secondary analysis is that the data set might be
deﬁcient or problematic in one or more ways, such as the sample used or the variables
measured (Polit & Hungler, 1995). However, the parent study’s sample - a convenience

sample of 102 women with a conﬁrmed diagnosis of breast cancer undergoing short-stay

23

surgery - provided a relatively large, random sample of women that represented the
population at risk for seroma formation. Moreover, the parent study deﬁned seroma
formation in the same way as the secondary analysis. Due to its experimental design, the
parent study provided a way to distinguish types of post-discharge care following breast
cancer surgery by offering a comparison between the parent study’s post-discharge
nursing intervention with conventional care, which included agency home care nursing,
care provided by friends or family, or by the patient herself (self-care).

Other variables in the secondary analysis, such as type of surgery, patient BMI,
and patient age, were easily obtained by combining existing variables from the parent
study. For example, patient BMI (BMI = weight in kilograms divided by height in meters
squared) was computed by converting patient height and weight into meters and
kilograms, and applying the appropriate formula. Patient age in years was derived by
subtracting patient birth date from surgery date. Type of surgery was reduced to two
categories by recoding modiﬁed radical mastectomy and wide excision with axillary
lymph node dissection as traditional surgery, and lumpectomy with axillary lymph node
dissection as breast-conserving surgery.

From an original sample of 115 subjects, 13 women who had a mastectomy
without axillary lymph node dissection were excluded from the secondary analysis.
Women who did not have axillary lymph node dissection did not experience signiﬁcant
lymphatic disruption, and therefore were not representative of the majority of breast
cancer patients. As a consequence, the secondary analysis is based on a sample of 102

women from the parent study.

24

Pr e s

A retrospective chart audit of 102 breast cancer patients participating in A Nursing
Intervention for Short-Stay Breast Cancer Surgery (DAMDl 7-96-1-6325) was
performed by a registered nurse on the research staff four months after breast cancer
surgery. These data were entered into the CI-3 interview database, and were reformatted
for secondary analysis in the SPSS 7.5 database.
Protection of Human Subjects

Approval ﬁom the University Committee on Research Involving Human Subjects
(UCRIHS) was obtained (see Appendix A). Secondary analysis falls under the l-E
exempt category, which is deﬁned as research involving the collection or study of
existing data, if the source is recorded by the investigator in such a manner that subjects
cannot be identiﬁed. There were no physical, psychological, social, legal or economic
risks associated with the secondary analysis. Data were blinded to maintain each
participant’s anonymity.
Qpprati pg Deﬁnitions of the Variables

The dependent variable, seroma formation, was coded as present or absent.
Absent was coded (0) and present was coded (1). Independent variables were deﬁned as
follows: Type of post-discharge care was coded as: (0) conventional care provided by
local home care agency nurses, family caregivers, friends, or the patient herself (control
arm), or (1) nursing care received by subjects in the intervention arm of the parent study.
Type of surgery was coded as: (0) breast-conserving surgery consisting of lumpectomy

with axillary node dissection; or (1) traditional surgery, which included modiﬁed radical

25

mastectomy with axillary node dissection or wide excision with axillary node dissection.
BMI was coded as a continuous two-digit number ranging from 18 to 83, and was
calculated by the formula resulting in weight in kilograms divided by height in meters
squared. Age was coded as a continuous three-digit number expressed as years ranging
from 021 to 100.
Data Management

Data was coded and entered into the SPSS 7.5 database for statistical analysis.
Frequency distributions and measures of central tendency were calculated for all variables
of interest. Data from the parent study was cleaned of all outliers and missing data.
Review of the literature shows that statistical techniques that test the association between
variables, such as t-tests and chi-square, were most often used. This secondary analysis
used logistic regression to consider the four independent variables as a collection of
factors that, in and of themselves, may be signiﬁcant, but in relation to the other variables
could be found to have signiﬁcant predictive value in regard to seroma formation.

Surgery has been established as an important factor in understanding seroma
formation. Studies of seroma formation have contained multiple risk factors which may
contribute to a better knowledge of how to predict the odds of developing a seroma when
particular risk factors are present. Since multiple risk factors may be involved, several
risk factors must be controlled for simultaneously in the analysis of variables associated
with seroma formation. Logistic regression uses maximum likelihood estimation to
analyze the relationship between multiple independent variables and a dependent variable

that is categorical. The types of regression problems for which logistic regression models

26

apply are characterized by the dependent variable being dichotomous or binary. In the
secondary analysis, the dependent variable is binary in the form of 1 or 0, present/absent.

The strength of logistic regression is that it requires far fewer assumptions about
the population in which the data originated. Nevertheless, it may produce misleading
results if ﬁndings are basically unstable, inconsistent, or idiosyncratic to a particular
sample. It maximizes chance associations in a data set and often produces ﬁndings that
vary substantially across samples, especially in studies using many predictor variables and
a small sample (Downs, 1988).

Randomization of the study subjects was successful with no differences found
between the intervention and control arms. Thomdike’s (1978) rule of thumb for a rough
estimate of the minimal sample size for multivariate analysis is 10 subjects per variable
plus 50 additional subjects. Secondary analysis of A Subacute Care Nursing Intervention
for Short-Stay Breast Cancer Surgery (DAMDl7-96-1-6325) met that criteria - it
examined four independent variables and had data from 102 subjects. The large sample in
the secondary analysis (n=102), allowed for a more accurate assessment of the coefﬁcient
estimate of the independent variables.

Logistic regression requires interval variables, that is, variables that can be
measured with numeric precision (Lewis-Beck, 1980). Non-interval variables can be
considered as dichotomies. In the secondary analysis, the dependent variable was a
dichotomy - seroma: present/absent. Two of the independent variables also were
dichotomous: (1) type of post-discharge care: intervention/control; and (2) type of

surgery: breast conserving/traditional. Patient BMI and age were interval variables.

27

The equation was: Log p/I - p = b0 + b, (type of post-discharge care) + b, (type
of surgery) + b 3 (BMD + b, (age in years) + e. Signiﬁcance of p was set at 0.05. The
hypothesis was that the risk for seroma formation increases with (1) conventional post-
discharge care (received by the parent study’s control arm), (2) traditional breast surgery,
(3) increasing BMI score, and (4) increasing patient age. Individual variables were
introduced simultaneously, beta coefficients were calculated, and the statistical
signiﬁcance of each coefﬁcient was determined.

Results
Demogpaphics

Participants included women age 21 or older admitted for short-stay surgery (48
hours or less) as ﬁrst treatment for breast cancer. One hundred and two women were
studied. Body Mass Index ranged from 18 to 83, with a mean of 30.45. Ages ranged from
33 to 84 years, with a mean of 57 years.

Of 102 women, 29 (28.4%) developed a seroma. A total of 78 (76.4%) women
elected to have breast conserving surgery, and 24 women (23.6%) had a traditional
modiﬁed radical mastectomy or wide excision with axillary lymph node dissection. There
was no statistically signiﬁcant difference in demographics between the intervention and
control arms of the parent study (see Table 1).

Demographics in Table 2 are taken from the parent study and are based on n =
100. Overall, 93% of participants were Caucasian, and 64% had at least some college
education. All participants spoke and read English, and all had received some level of

formal education. The majority of participants (58%) were employed before surgery.

28

TABLE 1

Fregmcy ef Type of Sggegy and Seroma Formation (n = 102)

 

Intervention (n=51) Control (n=51) Total (n=102)

 

 

 

n % n % n %
Type of Surgery
Breast Conserving 41 80.4% 37 71.5% 78 76.4%
Traditional 10 19.6% 14 28.5% 24 23.6%
Post-Surgical Complications
Seroma Formation 14 27.5% 15 29.4% 29 28.4%

 

Married women also comprised the majority (57%), compared to participants who were
divorced, single, or widowed (see Table 2).
Answers te Research Questions

Multiple regression is a technique used to predict a single dependent variable
using a number of independent variables. However, when the dependent variable is
dichotomous, as in the case of seroma formation (present/absent), a logistic function is
used to simulate a normal distribution. The maximum likelihood model selects the beta
coefficients that make the observed results of the dependent variable most likely, and
eliminates all independent variables that are not statistically signiﬁcant and therefore poor
predictors of the dependent variable (seroma formation).

A subset of independent variables that were considered potential predictors of the
dependent variable would make the observed rates of the dependent variable most
“likely”. As mentioned previously, the equation representing these variables is as follows:

Log p/I - p = b0 + b, (type of post-discharge care) + b2 (type of surgery) + b3 (BMI) + b4

29

TABLE 2

Demoggaphics (n = 100)

 

Intervention (n=51) Control (n=49) Total (n=100)

 

 

n % n % n %

Ethnicity

Caucasian 48 94.1% 45 91.8% 93 93.0%

Other 3 5.9% 4 8.2% 7 7.0%
Marital Status

Married 28 54.9% 29 59.2% 57 57.0%

Divorced/Separated 11 21.6% 7 14.3% 18 18.0%

Never Married 8 15.7% 3 6.1% 1 l 1 1.0%

Widowed 4 7.8% 10 20.4% 14 14.0%
Employment Status

Employed before surgery 26 51.0% 32 65.3% 58 58.0%

Not employed before surgery 25 49.0% 17 34.7% 42 42.0%
Education

Completed graduate degree 8 15.7% 12 24.5% 20 20.0%

Completed college 9 17.6% 6 12.2% 15 15.0%

Completed some college 16 31.4% 17 34.7% 33 33.0%

Completed high school 11 21.6% 9 18.4% 20 20.0%

Completed some high school 5 9.8% 3 6.1% 8 8.0%

Completed grade school 2 3.9% 2 4.1% 4 4.0%

No formal education 0 0.0% 0 0.0% 0 0.0%

 

 

(age in years) + e, where p = probability of seroma formation. The initial model for
predicting seroma formation included type of surgery, type of post-discharge care, BMI,
and age. Both BMI (p = 0.7925) and age (p = 0.2769) were found to be non-signiﬁcant in
predicting seroma formation and were eliminated from the model. Type of post-discharge
care was kept in the model regardless of its ability to predict seroma formation, as it was

an inherent part of the parent study design (see Table 3).

30

TABLE 3

Main Effect Medel (n = 102)

 

 

Variable 1.3. SE B 2
Type of Surgery 1.6575 0.5002 00008“
Type of Post-Discharge Care 0.0922 0.4732 0.8454

 

*p 5 0.001

To assess the magnitude of the relationship of type of surgery, type of post-
discharge care, BMI, and age to seroma formation, change in the maximum likelihood
was determined (see Table 4). In this case, -2 times the log of the likelihood (~2LL) was
used, since the likelihood is a small number (less than one). A good model is one that
results in a high likelihood of the observed results, which translates to a small value for
-2LL. For example, if a model ﬁts perfectly, the likelihood is 1, and -2 times the log
likelihood is 0.

The initial model includes the constant beta coefﬁcient only (Log p/I - p = b0+ e).
It is considered the baseline model, and had a -2LL of 121.111 This model was used for
comparison with the Full Model, which included all beta coefficients [b , (type of post-
discharge care), b2 (type of surgery), b 3 (BMI), and b 4 (age in years)]. The ~2LL decreased
to 108.552. The model chi-square, 12.56 (p = 0.014), is the difference between the initial
model and the Full Model. Thus, the model chi-square tests the null hypothesis, namely,
that the beta coefﬁcients for all of the terms in the current model, except the constant, are
0. The degrees of freedom for the model chi-square are the difference between the

ntunber of variables in the two models. The Full Model was found to be statistically

31

TABLE 4

Testing for Confounding Variables (n = 102)

 

-2LL Change in -2LL p

 

 

 

 

 

 

Baseline Model 121.111 ---- ----
b0 ---- ---- ---

Full Model 108.552 12.559 0.0140
b, Type of Post-Discharge Care ---- ---- 0.8426
b, Type of Surgery ---- ---- 0.0013
b, BMI ---- ---- 0.7925
b 4 Age ---- ----‘ 0.2769

Main Effect Model 110.198 1.678 0.0014“
b, Type of Post-Discharge Care --- ---- 0.8426
b, Type of Surgery ---- ---- 0.0013

*ps 0.01

signiﬁcant (p = 0.0014), therefore, it was necessary to determine which of the variables
was contributing to the success of the model.

Of the four independent variables - type of post-discharge care (deﬁned as the
parent study’s intervention or control group), type of surgery (traditional or breast-
conserving), BMI and age (in years) - only type of surgery was found to be statistically
signiﬁcant (p = 0.0013), indicating that type of surgery is a possible predictor of seroma
formation. None of the other variables in and of themselves contribute as a main effect
variable, and therefore are not considered to be predictors of seroma formation. As

mentioned earlier, type of post-discharge care must also remain in the model, not because

32

it was a main effect variable, but because it was the basis upon which the parent study
sample was randomized.

Type of surgery was identiﬁed as the main effect variable. However, the other
variables - type of post-discharge care, BMI and age - could be possible confounders or
effect modiﬁers on the main effect variable. To test for confounding effects, the question
becomes: Is there an increased risk of seroma formation for those recovering from
traditional surgery versus breast—conserving surgery, after adjusting for (1) type of post-
discharge care, (2) BMI, and (3) age. The baseline model now becomes the Main Effect
Model: type of surgery and type of post—discharge care, [b , (type of post-discharge care),
b2 (type of surgery) ] with a -2LL of 110.198.

To check for confounding effects of BMI and age on the Main Effect Model (the
new baseline model), a comparison was made between the -2LL of the full model [ b,
(type of post-discharge care) + b 2 (type of surgery)+ b 4 (age in years)], and the -2LL of
the Main Effect Model. The change in -2LL of the main effect was less than 20%[ X, =
1.678 (p z 0.05)], which is not statistically signiﬁcant. Therefore, age and BMI do not
confound the relationship between surgery and seroma formation.

In addition, the beta coefﬁcient of surgery in the baseline model was 1.6732. With
the addition of BMI and age, the beta coefﬁcient decreased to 1.6515. This comparison
presents with only a 0.9% change [p( b, (BMI) = 0.6746, and p(b, (age) = 06546)].
Therefore, neither BMI nor age confound the effect of surgery on seroma.

The ﬁnal step in analyzing the Main Effect Model was to evaluate whether type of

post-discharge care was an effect modiﬁer on type of surgery and seroma. In other words,

33

 

was the relationship between seroma formation and surgery type different for differing
types of post-discharge care? With p = 0.8426, post-discharge care was not found to be a
statistically signiﬁcant effect modiﬁer (see Table 4).

The -2LL for type of post-discharge care (110.236) and type of surgery (1 10.198)
were then compared, revealing a non-signiﬁcant difference of 0.038. This ﬁnding
indicated that type of postodischarge care does not confound the predictive effect that
surgery has on seroma formation.

The odds ratio for the main effect variable - type of surgery - was 5.3292
(p = .0008), which suggests that the risk of seroma formation may be over ﬁve times
greater for women receiving traditional surgery compared to women undergoing breast-
conserving surgery. This ﬁnding supports the hypothesis that women having traditional
surgery are at greater risk for seroma formation than women who opt for breast-
conserving surgery.

The Full Model included the remaining independent variables - patient BMI and
age - in order to assess whether or not these variables confounded the results of the Main
Effect Model. Log Likelihood (-2LL) results were compared between the two models to
determine if the new variables confounded the relationship between surgery and seroma
formation. Comparison of beta coefﬁcients revealed a change of 0.9% between the Full
Model and the Main Effect Model, which indicates that neither BMI nor age confound the
effect of surgery on seroma formation.

Effect modiﬁcation was used to test whether the relationship between seroma

formation and surgery type was modiﬁed by the combination of type of surgery and BMI .

34

If BMI was an effect modiﬁer, then the relationship between seroma formation and
surgery type would be different for different levels of age and BMI. Results revealed p =
0.6746, and suggested that the combination of age and surgery had no effect on the
relationship between surgery and seroma formation. Similarly, the modifying effect of
type of surgery and age showed that p = 0.6546. Again, the conclusion is that the
combination of BMI and type of surgery did not signiﬁcantly modify the relationship
between surgery and seroma formation.

Discussion
Interpretation of the Findings

The hypothesis that there is an increased risk of seroma formation with (1) post-
discharge care received by the parent study’s control arm, (2) patient BMI and (3) patient
age was not substantiated. The only hypothesis found to predict seroma formation was
related to type of surgery. Results of the logistic regression indicate that women having
traditional surgical procedures that involve the removal of large amounts of breast tissue
may have a ﬁve times greater risk of developing a postoperative seroma compared to
women undergoing breast-conserving surgery.

Variables such as patient age and weight have been found to be associated with
seroma formation in previous studies that utilized parametric tests with large samples,
and non-parametric tests when sample sizes were small. This secondary analysis
attempted to reproduce those ﬁndings with a more powerful statistical method - logistic
regression. However, considering the contradictory ﬁndings represented in the literature,

it is not surprising that the demographic variables failed to predict seroma formation

35

 

under more rigorous testing. Nevertheless, the impact of demographic variables may be
misrepresented in this secondary analysis, given that 76.4% of the participants elected to
have breast conserving surgery, compared to only 23 .6% who had mastectomy or wide
excision with axillary lymph node dissection. It is possible that these factors do have an
impact on seroma formation for women undergoing traditional surgery, as previous
studies have suggested, but were not detectable in this study because of the small
sampling of women having traditional surgery. Perhaps a study with equal group size
would yield different results.

Results of the secondary analysis suggest that type of post-discharge care does not
inﬂuence the risk of postoperative seroma formation. Given the fact that the etiology of
seroma formation remains unknown, the preventive effects of post-discharge nursing care
cannot be determined. Therefore, the focus of post-discharge nursing should be on
preventing the serious sequella of seroma formation through patient education, early
detection, and prompt treatment.

In terms of type of surgery and its effect on seroma formation, the recent shift to
less invasive, short-stay surgery has left the literature wanting. Few studies compare
traditional surgical techniques with breast-conserving surgery. Nevertheless, results of
this secondary analysis suggest that breast-conserving techniques may result in
signiﬁcantly fewer postoperative seromas than traditional breast surgery.

Diseyesiop of Results with the Conceptual Model
With regard to Orem’s conceptual model, the results of this secondary analysis

indicate that self-care may commence before surgery, when the patient begins to consider

36

surgical options. Breast-conserving surgery is relatively new, and the consequences of
this technique are not well known. Advanced Practice Nurses who provide patient
education before surgery can incorporate the ﬁndings about increased risk of seroma
formation with traditional surgery into their teaching about the consequences of surgical
intervention, which may help women decide among surgical treatment options.

The secondary analysis indicated that the type of post-discharge care women
received after breast cancer surgery had no effect on the incidence of seroma. As a
consequence, Orem’s model of self-care cannot be fully applied to the problem of seroma
formation. Orem’s focus is on learned, goal-oriented actions performed by the patient or
the nurse in order to meet therapeutic demands deﬁned by a particular health deviation.
Since the secondary analysis indicated that the type of post-discharge care did not have an
effect on seroma formation, no new information was revealed to suggest that particular
actions taken by the APN prevent seroma formation. Thus, the aim of self-care after
surgery is early detection of seroma formation, and prompt medical treatment in order to
prevent infection, semis, and delayed wound healing.

Implications for the Advanced Practice Nurse

In an attempt to reduce health care costs, surgical management of breast cancer
patients now consists of hospital stays of 24 hours or less, and patients are increasingly
expected to provide their own care during the recovery period at home. The APN is
uniquely qualiﬁed to design patient education that emphasizes self-care, and to provide
supervision of post-discharge progress in the home. In order to provide continuity of care

and to ensure positive post-surgical outcomes for breast cancer patients, utilization of the

37

entire interdisciplinary health team may be the most effective approach to achieving
comprehensive patient care.

For the APN, Orem’s model provides a practical framework from which to plan
pre and post-discharge education for the breast cancer patient. In light of the ﬁndings of
the secondary analysis, pre-surgical education should include information about the
various surgical options, including traditional and breast-conserving methods, and the
effects of a particular surgical method on seroma formation. After surgery, patient
education should include information about the likelihood and consequences of seroma
formation, the signs and symptoms of seroma, and a plan to seek prompt medical
treatment if seroma formation occurs.

Case management of the breast cancer patient is increasingly important as time
spent in the tertiary health care setting decreases. Outcome goals which include the
prevention of serious sequella of seroma formation need to be clearly identiﬁed,
discussed with the patient, and carefully monitored in order to ensure patient progress and
to improve standards of care.

Recommen ti for urther esearch

Much remains to be learned about the etiology of seroma formation. Although the
prevalence of seroma is considerable, its prevention is not yet feasible. Consequently,
more attention should be paid to preventing negative outcomes of seroma. Nursing
studies that document how prompt recognition and treatment of seroma reduce serious
complications, such as infection and delays in wound healing, could be used to improve

post-discharge planning and lower health care costs.

38

Results of this secondary analysis indicate that seroma formation affects 28.4% of
post-surgical breast cancer patients. Variables not considered in the secondary analysis,
such as medical comorbidities and immune system characteristics, may shed light on the
cause of seroma formation. Therefore, continued research into the etiology and
management of seroma formation is needed to improve post-surgical outcomes for breast
cancer patients.

Summm

Secondary analysis of 102 participants from A Subacute Care Nursing
Intervention for Short-Stay Breast Cancer Surgery (DAMDl 7-96-1-6325) suggests that
seroma formation affects a large number of women following breast cancer surgery.
Based on the unequal group size of this analysis, patients undergoing traditional surgery
(modiﬁed radical mastectomy or wide excision with axillary lymph node dissection) may
have a ﬁve-times greater risk of seroma formation, compared to patients opting for
breast-conserving surgery, regardless of the type of post-discharge care they received, or
personal characteristics such as BMI and age.

The etiology of seroma formation is unknown, and its prevention remains
problematic. Before surgery, the APN on the interdisciplinary health team should instruct
patients on making the best surgical decision for their set of circumstances. After surgery,
the APN can help prevent the serious sequella of seroma formation by teaching the
patient how to recognize the signs and symptoms of seroma formation, and by developing

a plan to seek prompt medical treatment if seroma formation occurs.

39

APPENDIX

40

APPENDIX A

The University Committee on Research Involving Human Subjects IRB# 99154

41

 

OFFICE OF
RESEARCH
AND
GRADUATE
STUDIES

liversity Committee on
Research Involving
Human Subjects
(UCRIHS)

Michigan State Univeisny
6 Administration Budding
East Lansmg. Michigan
48824-1046

517/355-2180
FAX 517/353-2976

tricniaan 51.213 universnv
55-2 .5 ifle/‘lUIlO/IJI diversity
Fuse/lance in Action

   
  

MICHIGAN STATE

U N I V E R S IT Y
March12,1999

 

TO: Dr. Gwen WYATT
A230 Life Sciences
College Of Nursing

RE: IRB# 99154 CATEGORY: 1-E
APPROVAL DATEzMaI'ch 9, 1999

TITLEzPOSTOPERATIVE SEROMA FORMATION FOLLOWING BREAST CANCER
SURGERY: IMPLICATIONS FOR THE ADVANCED PRACTICE NURSE

The University Committee on Research Involving Human Subjects' (UCRIHS) review of
this project is complete and I am pleased to advise that the rights and welfare of the
human subjects appear to be adequately protected and methods to obtain informed
consent are appropriate. Therefore, the UCRIHS approved this project.

RENEWALS: UCRIHS approval is valid for one calendar year, beginning with the
approval date shown above. Projects continuing beyond one year must be renewed
with the green renewal form. A maximum of four such expedited i'enewals possible.
Investigators wishing to continue a project beyond that time need to submit it again for a
complete review.

REVISIONS: UCRIHS must review any changes in procedures involving human
subjects, prior to initiation of the change. If this is done at the time of renewal, please
use the green renewal form. To revise an approved protocol at any other time during
the year, send your written request to the UCRIHS Chair, requesting revised approval
and referencing the project's IRB# and title. Include in your request a description of the
change and any revised instruments, consent forms or advertisements that are
applicable.

PROBLEMS/CHANGES: Should either of the following arise during the course of the
work. notify UCRIHS promptly: 1) problems (unexpected side effects. complaints, etc.)
involving human subjects or 2) changes in the research environment or new information
indicating greater risk to the human subjects than existed when the protocol was
previously reviewed and approved.

If we can be of further assistance, please contact us at 517 355-2180 or via email:
UCRIHS@pilot.msu.edu. Please note that all UCRIHS forms are located on the web:
http://www.msu.edu/unit/vprgs/UCRIHS/

’ cerel

 

, RIHS Chair
DEW: ’
cc ary Bloomfield

42

REFERENCES

43

REFERENCES
American Cancer Society (1996). Exercises after breast surgepy [Brochure].
Atlanta, GA: Author.

Bloch, S. & Kissane, D. (1995). Psychosocial care and breast cancer. The Lancet

 

34_6, 1114.
Burak, W., Goodman, P., Young, D., & Farrar, W. (1997). Seroma formation
following axillary dissection for breast cancer: Risk factors and lack of inﬂuence of

bovine thrombin. Jopmal of Sggigl Oncolegy, 64, 27-31.
Clark, J. & Kent, R. (1992). One-day hosptialization following modiﬁed radical

mastectomy. American Surgeon, 58, 239-242.

Chilson, T., Chan, F., Lonser, R., Wu, T., & Aitken, D. (1992). Seroma
prevention after modiﬁed radical mastectomy. American Surgeon 12, 750-754.

Dietrick-Gallagher, M. & Hyzinski, M. (1989). Teaching patients to care for
drains after breast cancer surgery. Oneology Nursing Forum, 16, 263-265.

Downs, F. (1988). Handbook of Research Methodology. New York: American
Journal of Nursing Company.

Fisher, B., Redmond, C., & Poisson, R., (1989). Eight year results of a
randomized clinical trial comparing total mastectomy and lumpectomy with or without
irradiation in the treatment of breast cancer. New England Jopmﬂ of Medicine, 320,
822-828.

Goodman, A. & Mendez, A. (1993). Deﬁnitive surgery for breast cancer

performed on an outpatient basis. Archives of Surgegz, 128, 1149-1152.

44

Holcombe, C., West, N., Mansel, R. & Horgan, K. (1995). The satisfaction and
savings of early discharge with drain in situ following axillary lymphadenectomy in the
treatment of breast cancer. European Journal of Surgical Oncology, 21, 604-606.

Kambouris, A. (1995). Physical, Psychological, and economic advantages of
accelerated discharge after surgical treatment for breast cancer. The American Surgeon,
62, 123-127.

Kumar, S., Lal, B., & Misra, M. (1995). Post-mastectomy seroma: A new look
into the aetiology of an old problem. The Journal of The Royal College of Surgeons of
Edinburgh, 40, 292-294.

Lewis-Beck, M. (1980). Applied Reggession: An Introduction. Beverly Hills: Sage

McManus, S., Topp, D., & Hopkins, C. (1994). Advantages of outpatient breast

surgery. American Surgeon, 60, 967-970.

Morrow, M. & Hassett, C. (1994). Breast cancer in young women: Issues in local
therapy. Journal of the National Cancer Institute Monographs, 16, 79-84.

Parikh, H., Badwe, R., Ash, C., Hamed, H. Freitas, R., Chaudary, M. & Fentirnan,
I. (1992). Early drain removal following modiﬁed radical mastectomy: A randomized
trial. Jogml of Surgical Oncology, 51, 266-269.

Petrek, J ., Peters, M., Cirrincione, C., & Thaler, H. (1992). A prospective
randomized trial of single versus multiple drains in the axilla after lymphadendectomy.
Surgery, Gypecolegy & Obstetrics, 175, 405—409.

Polit D. & Hungler, B. (1995). Nursing Research: Principles and Methods (5‘'1

ed.). Philadelphia: Lippincott

45

Sciartelli, C. (1995). Using a clinical pathway approach to document patient
teaching for breast cancer surgical procedures. Oncology Nursing Forum, 22, 131-137.

Somers, R., Jablon, L. Kaplan, M., Sandler, G., & Rosenblatt, N. (1991). The use
- of closed suction drainage after lumpectomy and axillary node dissection for breast
cancer. Annals of Surgery, 215, 146-149.

Tejler, G. & Aspergren, K. (1985). Complications and hospital stay after surgery
for breast cancer: A prospective study of 385 patients. British Journal of Surgeg, 72,
542-544.

Thorndike, R. (1978). Correlational Procedures for Research. New York: Gardner
Press, Inc.

van Wersch, A., Bonnema, J ., Prinsen, B., Pruyn, J. Wiggers, T., & van Geel, A.
(1997). Continuity of information for breast cancer patients: the development, use and
evaluation of a multidisciplinary care-protocol. Patient Education and Counseling, 30,
175-1 86.

Veronesi, U., Banﬁ, A., & Del Vecchio, M. (1986). Comparison of Halsted
mastectomy with quandrantectomy, axillary dissection, and radiotherapy in early breast
cancer. Eurogan Journal of Cancer, 22, 1085-1089.

Wardlaw G. & Insel, P. (1993). Persxctives in Nutrition (2"d ed.). St. Louis:
Mosby.

Wyatt, G. (Principle Investigator), Given C. & Given B. (Co-Principle
Investigators). (Submitted 9/15/95). A Subacute Care Intervention for Short-Stay Breast

Cancer Surgeg. Funded 9/15/96 by the Department of Defense, grant #DAMD17-96-6325.

46

Yii, M., Murphy, C., & Orr, N. (1995). Early removal of drains and discharge of

breast cancer surgery patients: A controlled prospective clinical trial. Annals of the Royal

College pf Sggery of England, 77, 377-379

47

HICH

N TRT

169 S E UNIV. LIBRARIES

312 3018354005