Feasibility of Implementing a Breast Reconstruction Database

Abstract

Objective:

To assess whether implementing a breast reconstruction database would be feasible in terms of time commitment, cost, and overall benefits in a tertiary-care hospital.

Methods:

A survey was sent to 40 Canadian plastic surgeons who have a practice focused on breast reconstruction. The survey assessed demographics, practice characteristics, database use, and opinions on database construction. Univariate descriptive analyses were performed on all variables.

Results:

Thirty-one surgeons responded to the survey (77.5%). Most were from Ontario (29.1%) and worked in an academic center (83.9%). Of all, 45.3% of surgeons performed more than 50 breast reconstructions yearly. Six (19.4%) surgeons utilized databases that were all started for quality improvement and research purposes. Databases included variables such as demographics, type of reconstruction, complications, surgeons involved, and type of implants. Data are input by research assistants (50%) for approximately 4.2 hours per month at a cost below 200$CAD per month. Databases are funded by research grants (50%), hospital funds (33.3%), and/or division funds (16.7%). Of the surgeons without databases, 60% have considered starting a database. Barriers include being too busy (72%) and impressions of the cost being too high (32%). Surgeons commonly felt that a database would be beneficial at their practice (80%), provincially (77.4%), and nationally (67.7%).

Conclusions:

Plastic surgeons are open to the idea of constructing a breast reconstruction database and that the costs and time required are lower than expected. Grants or integration with existing databases should be pursued on a provincial level first prior to pursuing a national database.

Introduction

Breast cancer is the most common type of cancer and second most common cause of death among Canadian women.¹ In 2016, it is estimated that 25 700 women were diagnosed with breast cancer in Canada (130.1 cases per 100 000 women) and that approximately 4900 will die from the disease. The 5-year survival for women diagnosed with breast cancer is 87%.¹ While the incidence of breast cancer in Canada has been decreasing, the disease is still prevalent and causes substantial physical and emotional harm.¹

Surgery is one of the crucial elements of breast cancer treatment, and women diagnosed with breast cancer may have the option of either a breast-conserving procedure or a mastectomy. Approximately 39% of Canadian women diagnosed with breast cancer undergo a unilateral or bilateral mastectomy.² The incidence of mastectomy varies considerably between provinces, with a low of 26% in Quebec to a high of 69% in Newfoundland. A number of factors may explain this interprovincial variation, including longer travel times to radiation facilities, provincial age variations, and socioeconomic factors.² Breast reconstruction is associated with high patient satisfaction and improved psychosocial outcomes.³ Rates of breast reconstruction among Canadian women are historically low compared to other developed countries.³ Research has not explored the reasons for the lower rates in Canada. Negative predictors of undergoing breast reconstruction are increasing age, nonwhite ethnic background, low income, and nonurban residence.⁴

Databases capturing breast reconstruction are crucial to determining demographics, utilization rates, predictors of prognosis, and adverse events. Most epidemiological studies use large administrative databases to capture the incidence of breast reconstruction following mastectomy, which have a number of limitations. The databases may not accurately capture important variables such as smoking, obesity, comorbidities, and socioeconomic factors.⁵ In addition, large administrative databases do not typically capture outpatient reconstructive surgeries completed after the initial mastectomy.³

The goal of this study is to assess whether constructing and implementing a breast reconstruction database is feasible in terms of time commitment, cost, and benefits in a tertiary-care facility. It is hypothesized that a breast reconstruction database is too expensive and time consuming to be feasible in one’s surgical practice.

Methods

A web-based survey was distributed to a convenience sample of 40 plastic surgeons across Canada who have a surgical practice focused on breast reconstruction. The survey comprised both open- and closed-ended questions related to demographics, practice characteristics, breast reconstruction metrics, database use, variables tracked, and feasibility of implementing a breast reconstruction database. A total of 2 reminders were sent to participants to complete the survey. Institutional ethics approval was obtained for this study. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008.⁵ Informed consent was obtained from all patients for being included in the study

Univariate descriptive analyses were performed for all variables. The survey was distributed using Opinio Survey Software Version 7.9 (Object Planet, Oslo, Norway). SPSS version 24 was used for all analyses (IBM SPSS Statistics, Armonk, New York).

Results

A total of 31 plastic surgeons completed the survey (77.5% response rate; Table 1). Most surgeons are from Ontario (29.1%), Alberta (22.6%), and Quebec (12.9%) and have been practicing for an average of 13.5 years (Standard deviation [SD]: 9.2 years). Their practices are mostly fee for service (90.3%), in an academic health sciences center (83.9%), and solo (41.9%) or hospital based (41.9%). The surgeons work an average of 54.7 hours per week (SD: 9.7) and are responsible for an average of 5.4 (SD: 2.1) call shifts per month. The number of breast reconstruction performed yearly varies, with 38.7% performing between 26 and 50 per year and 45.3% performing more than 50 per year. Implant reconstructions (57.2%) and autologous tissue only (20.6%) represent the majority of procedures (Table 1).

Table 1.

Demographic Characteristics and Breast Reconstruction Metrics of Plastic Surgeons That Completed the Breast Reconstruction Survey.^a

Variable	Frequency (%)
N	31 (77.5)
Province of practice
Ontario	9 (29.1)
Alberta	7 (22.6)
Quebec	4 (12.9)
Manitoba	3 (9.7)
Saskatchewan	3 (9.7)
British Columbia	2 (6.5)
Newfoundland	1 (3.2)
Nova Scotia	1 (3.2)
New Brunswick	1 (3.2)
Years practicing
0-5	9 (29.1)
6-10	6 (19.4)
11-15	3 (9.7)
16-20	5 (16.1)
>20	8 (25.8)
Mean (SD)	13.5 (9.2)
Fee structure of practice
Predominantly fee for service	28 (90.3)
Predominantly salary	1 (3.2)
Blended	2 (6.5)
Main work setting
Academic health sciences centre	26 (83.9)
Non-academic health science teaching hospital	3 (9.7)
Private office/clinic	1 (3.2)
Community hospital	1 (3.2)
Type of practice
Solo	13 (41.9)
Hospital based	13 (41.9)
Group	5 (16.2)
Hours worked per week
30-40	3 (10)
41-50	13 (41.9)
51-60	10 (32.3)
>60	5 (16.1)
Mean (SD)	54.7 (9.7)
Call shifts per month
0-5	22 (71)
6-10	8 (25.8)
>10	1 (3.2)
Mean (SD)	5.4 (2.1)
Number of breast reconstructions performed yearly
0-25	5 (16.2)
26-50	12 (38.7)
51-75	5 (16.2)
76-100	5 (16.2)
>100	4 (12.9)
Mean percentage of procedure type (SD)
Implant reconstruction	57.2 (22.5)
Autologous tissue only	29.5 (20.6)
Autologous and implant (latissimus dorsi)	15.7 (13.8)
Other	5 (0)

^a N = 31.

A total of 6 (19.4%) plastic surgeons from our sample currently utilize a breast reconstruction database (Table 2). Excel (33.3%) and Synoptec (33.3%) are most commonly used, the networks are typically externally hosted (66.7%), and surgeons feel their software is easy to use (100%). Databases were started for both quality improvement and research purposes (100%), and variables tracked include demographics (100%), type of reconstruction (100%), complications (100%), unilateral versus bilateral (100%), surgeons involved (100%), type of implants (83.3%), and many others (see Table 2). Surgeons would like their database to also include photos (33.3%), complications (16.7%), quality-of-life data (16.7%), and patient costs (16.7%). Data are mostly input by research assistants (50%) and the surgeon (33.3%), for an average of 4.2 hours per month (SD: 3.4) at a cost of between $100 and $200 (50%). The databases are funded primarily by research grants (50%), hospital funds (33.3%), and division funds (16.7%). Most databases receive input from fewer than 5 (83.3%) surgeons, and the use is well established in their respective practice (83.3%; Table 2).

Table 2.

Database use variables of plastic surgeons that currently utilize a breast reconstruction database.^a

Variable	Frequency (%)
Do you currently track your breast reconstructions using a database?
Yes	6 (19.4)
No	25 (80.6)
Type of database used
Excel	2 (33.3)
Synoptec	2 (33.3)
SPSS	1 (16.7)
HealthScreen	1 (16.7)
Location of network
Internally hosted	2 (33.3)
Externally hosted	4 (66.7)
Do you feel that the database software is easy to use?
Yes	6 (100)
No	0
Are you (and other surgeons) satisfied with the current database?
Yes	5 (83.3)
Unsure	1 (16.7)
Why did you initially establish the database?
Both quality improvement and research	6 (100)
Other (list for volunteers and support access)	1 (16.7)
What variables do you track?
Demographics	6 (100)
Type of reconstruction	6 (100)
Complications	6 (100)
Unilateral versus bilateral	6 (100)
Type of mastectomy	6 (100)
Surgeons’ involved	6 (100)
Type of implants	5 (83.3)
Recipient vessels	5 (83.3)
Pictures	4 (66.7)
Follow-up dates	3 (50)
Single surgeon versus team approach	2 (33.3)
Number of surgeries	1 (16.7)
Visits per patient	1 (16.7)
Are there other functions you would like your current software to provide?
Photos	2 (33.3)
Complications	1 (16.7)
Quality of life data	1 (16.7)
Patient costs	1 (16.7)
Who inputs the data?
Research assistant	3 (50)
Surgeon	2 (33.3)
Secretary	1 (16.7)
Hours per month inputting data
0-5	4 (66.7)
6-10	2 (33.3)
Mean (SD)	4.2 (3.4)
Cost per month to maintain database
100-200	3 (50)
No cost	2 (33.3)
Unknown	1 (16.7)
How is the database funded (multiple sources)?
Research grants	3 (50)
Hospital funds	2 (33.3)
Division/departmental funds	1 (16.7)
Surgeons salary	3 (50)
How is your software subscription paid?
Unsure	4 (66.7)
Annually	2 (33.3)
How many surgeons input data into the database?
1	3 (50)
2-5	2 (33.3)
>5	1 (16.7)
Is the regular use of the database well established in your practice?
Yes	5 (83.3)
No	1 (16.7)
Are there in house policies promoting database use?
Yes	2 (33.3)
No	3 (50)
Unsure	1 (16.7)

^a N = 31.

For those plastic surgeons that do not currently have a database, approximately half have considered starting one although they have not heard of specific software (60%; Table 3). Barriers to starting a database include being too busy (72%), too costly (32%), not interested (16%), too few patients (8%), no utility (8%), no research supporting their use (4%), and not enough support (4%). Most surgeons feel that a database would be beneficial at their practice (80%) and that provincial (77.4%) and national (67.7%) databases would be useful (Table 3).

Table 3.

Variables Related to the Potential Creation of a Breast Reconstruction Database.

Variable	Frequency (%)
If you do not use a database, have you considered starting one?
Yes	12 (48)
No	9 (36)
Unsure	4 (16)
If so, what software have you considered?
None	15 (60)
Missing	10 (40)
If you do not use a database, what are some barriers to starting one?
Too busy	18 (72)
Too costly	8 (32)
Not interested	4 (16)
Too few patients	2 (8)
No utility	2 (8)
No research supporting its use	1 (4)
Not enough support	1 (4)
Do you feel a database at your practice would be beneficial?
Yes	20 (80)
No	1 (4)
Unsure	4 (16)
Do you feel that a provincial database would be beneficial (all respondents)?
Yes	24 (77.4)
No	1 (3.2)
Unsure	6 (19.4)
Do you feel that a national database would be beneficial (all respondents)?
Yes	21 (67.7)
No	1 (3.2)
Unsure	9 (29.1)

^a N = 31.

Discussion

The goal of this study was to determine whether starting a breast reconstruction database is feasible in terms of time commitment, cost, and benefits by surveying plastic surgeons across Canada who focus on breast reconstruction. Our survey had a high response rate of 77.5%, with the majority of surgeons coming from central Canada, working in an academic health sciences center, and approximately half performing more than 50 breast reconstructions per year. Few plastic surgeons currently utilize a breast reconstruction database; however, those who do track a number of useful variables and do so at a low cost and with minimal hours per month needed to maintain the database. Approximately half of those who do not currently have a database have considered starting one, with the main barriers being too busy or perceived cost being too high. The survey indicated that plastic surgeons believe databases at their practice, provincially, and nationally would be useful and however would likely be difficult to implement.

The demographics of our sample are comparable to those found in other studies.^6-8 Research has shown that the majority of plastic surgeons work in urban centers, at teaching hospitals, and frequently belong to hospital-based or solo practices. Plastic surgeons in general have been shown to work an average of 50 to 55 hours per week, while the number of call shifts per month varies considerably depending on location, subspecialty, and type of practice.^6,7 While a number of studies have evaluated the incidence of breast reconstruction in Canada and the United States,^3,8-11 we were unable to find studies investigating the number of reconstructions a surgeon typically performs on a yearly basis. The aim of our study was to distribute the survey to surgeons who focus on breast reconstruction, and our survey reflects this goal as only 5 of the 40 surgeons who responded to the survey performed less than 25 breast reconstructions per year. Our results agree with the literature in terms of the frequency of procedure type, as implant reconstruction followed by autologous tissue and latissiumus dorsi are most commonly used in Canada and the United States.^12-14

There are numerous benefits to creating a comprehensive breast reconstruction database, including producing a better understanding of practice characteristics, diseases, the range of treatments, and both short- and long-term outcomes.^15,6,7 Complex databases can be used as resources for the generation of research hypotheses and the completion of future studies. The American Society of Plastic Surgeons believes that it is key that plastic surgeons have an effective mechanism for collecting clinically relevant information from patients over a longitudinal period of time and benchmark their patient outcomes and complications data against other groups of patients.¹⁵ With the usefulness of registries depending on the quality of the data, it is crucial that data be complete, contain the appropriate variables, and be validated on a regular basis to ensure accuracy.

A review of the literature indicates that there are currently no studies, to our knowledge, assessing the opinions of plastic surgeons on constructing a breast reconstruction database or the characteristics of databases currently in use. The incidence and demographic variables of women undergoing breast reconstruction in Canada are typically reported using data from the Hospital Morbidity Database and National Ambulatory Care Reporting System, which are maintained by the Canadian Institute for Health Information. This is a large database that includes a variety of validated demographic and administrative variables but lacks variables specific to breast reconstruction and other comorbidities that may affect initial surgical treatment and prognosis. Databases in the United States include the American Society of Plastic Surgeon’s Tracking Operations and Outcomes for Plastic Surgeons database, the Surveillance, Epidemiology and End Results database, and the National Cancer Data Base. These databases track a number of important variables including risk factors, procedural and 30-day outcome data, and tumor characteristics.^3,15,16 However, they do not capture reconstructions performed outside the initial mastectomy, patient-reported outcomes, and long-term outcome data that can be of considerable use for quality improvement initiatives and tracking complications.

Studies concerning medical databases in general have shown that tracking data can be of considerable use in terms of improving outcomes, identifying risk factors, understanding the physical and psychosocial impact of certain procedures, and enhancing the design of interventions.^16,17-20 If we extrapolate based on our results, only a handful of plastic surgeons in Canada are likely to be tracking their breast reconstruction patients using a database. This represents a sizable data drain because studies concerning risk factors and outcomes, for example, are difficult and time consuming without a proper database. Our results indicate that the construction and maintenance of a database is likely much easier and cheaper than initially suspected, with informal programs such as Excel being used frequently and with low cost. The main argument for any sort of medical intervention or quality improvement initiative is typically the time required to ensure its continuous and proper usage. We acknowledge that this is a substantial limitation and one that could likely be addressed with substantial support through funding initiatives, integration with other provincial or national databases, and creating internal policies surrounding the use of a database.

In order to translate our findings into real-world application, provincial- or national-level grants may be pursued. Grants from the Canadian Society of Plastic Surgeons or the Canadian Institute of Health Research, among other funding organizations, could provide necessary funds to implement such a database. Other ideas for implementing a database include adding breast reconstruction variables to existing oncologic databases or electronic medical records, which would eliminate the difficult and time consuming initial stages of database construction and validation. For example, the province of Alberta currently has a database where a good portion of cancer and breast reconstruction surgeries are captured and analyzed. If this avenue is pursued, then care should be taken to include all relevant variables and outcome data specific to breast reconstruction such as type of reconstruction, complications, type of implant, photos, quality-of-life data, patient costs, and more. While a national database would be ideal, health care is a provincial mandate, so it is likely more practical to approach this from a provincial standpoint first and then attempt to construct a national database. In addition, focusing on provincial databases first means that provinces with databases can help leverage other provinces to implement the same and learn from their mistakes. In order for a database to succeed, policies surrounding database usage should also be established through internal means first then provincially should the time come.

The main strengths of this study include investigating an area with sparse research, utilizing a sample size that is highly generalizable, and the potential for prompt knowledge translation of the results. The limitations of the study include the inability to survey all plastic surgeons across Canada, response bias of included plastic surgeons, and the fact that few plastic surgeons utilize a database thereby limiting the potential generalizability of our recommendations for database construction.

Conclusions

We have presented the results of a survey assessing whether starting a breast reconstruction database would be possible in terms of time commitment, cost, and benefits. Plastic surgeons across Canada believe that a database would have considerable use at all levels of health care, although are concerned that the time and costs of maintaining such a database would be too great. Our responses indicate that this may not be the case, given the administrative characteristics of databases currently in use. At this time, funding or integration with other electronic databases may be pursued if possible.