Assessing First and Multiple Reoperations in 23,301 Breast Reconstructions: Immediate Versus Delayed Reconstructions in Women with Breast Cancer

Simple Summary

We performed one of the largest recent assessments of risk of reoperation by reconstruction timing while accounting for radiotherapy and reconstruction techniques. In this cohort study, 23,301 primary mastectomies in women with breast cancer undergoing either immediate breast reconstruction (IBR) or delayed reconstruction were identified between 2010 and 2022. We found that the first reoperation incidence and risk were higher in IBR patients compared to those who delayed reconstruction, although we failed to detect a difference for multiple returns to surgery, except in certain subgroups. Preoperative expectations play a significant role in patient satisfaction and quality of life. Assessing reoperation risk by timing among different reconstruction modalities can aid patients in making informed decisions about the type of breast reconstruction to undergo.

Abstract

Background: Few studies have compared the risk of reoperation by timing in breast reconstruction surgery after mastectomy. We evaluated the first and total number of reoperations by reconstruction timing in women with breast cancer undergoing primary mastectomy. Methods: A cohort study of 23,301 primary mastectomies in women with breast cancer undergoing either immediate breast reconstruction (IBR) or delayed reconstruction was carried out within Kaiser Permanente between 2010 and 2022. The first reoperation rate was calculated using cause-specific Cox Proportional Hazards Models, while Multiplicative Cox Proportional Hazards Models were used to account for mortality and timing in reoperation. Patients were continuously monitored for death, outcome of interest, loss to follow-up through healthcare membership termination, or study end date (31 December 2022). Results: In total, 78.4% (n = 18,276) of the cohort underwent IBR. The average follow-up time was 5.9 years (±3.8). The following covariates were imbalanced (standardized mean difference [SMD] ≥ 0.20) between IBR and delayed groups: BMI, smoking status, year of mastectomy, bilateral procedures, and reconstruction type. The crude incidence of first reoperation was 33.04% vs. 31.72% in IBR vs. delayed patients and the risk of reoperation was 18% higher in IBR patients (HR = 1.18, 95% CI = 1.12–1.25). There was no difference in the risk of reoperation by timing (p > 0.05) when assessing multiple reoperations. The reoperation risk was the highest for IBR patients who did not complete reconstruction or single-stage reconstruction. In addition, the first reoperation rate of IBR patients was higher in those who underwent expander–implant-based reconstruction. Conclusions: The first reoperation rate was higher in IBR patients compared to those who delayed reconstruction, although we failed to detect a difference for multiple returns to surgery, except in certain subgroups. Assessing reoperation risk by timing among different reconstruction modalities can aid patients in making informed decisions about the type of breast reconstruction to undergo.

1. Introduction

Breast cancer is the most common cancer in women in the United States, except for non-melanoma skin cancers, and the American Cancer Society projects that 316,950 new cases of invasive breast cancer will be diagnosed in women in the United States in 2025 [1]. The decision and timing of when to conduct breast reconstruction can be controversial and often depends most on individual circumstances and whether adjunctive radiotherapy is required [2].

Immediate breast reconstruction (IBR) is conducted at the same time as mastectomy, while delayed reconstruction may be performed at any time following mastectomy. Acceptability of immediate reconstruction within the surgical community has evolved over time [3,4]. While plastic surgeons provide anecdotal and evidence-based clinical information and advice, patient decisions regarding breast reconstruction are very personal and are influenced by individual values and preferences [5]. Some women may place more value on shorter surgical procedures, less recovery time, and fewer scars; others may value a lower long-term follow-up burden, which may be the result of a lengthier surgical procedure and more painful recovery [6,7]. An increasing number of women are choosing to undergo IBR as it has been shown to lessen the effects on self-image and psychosocial well-being observed following mastectomy compared with delayed reconstruction [8]. While IBR has been shown to be safe [9,10,11,12,13], there have been few studies evaluating the risk of multiple surgeries by breast reconstruction timing [14], particularly with regard to reconstruction choices including laterality or implant usage [15,16,17,18]. This can create additional challenges for clinicians as well as increase medical expenses [19,20]. Most importantly, it places a burden on patients as the number of follow-up surgeries and the time required to complete their breast reconstruction do not always meet their expectations.

In this study, we sought to compare the risk of the first and total number of returns to the operating room (OR) by reconstruction timing in women with breast cancer undergoing primary mastectomy in a universal coverage system across multiple US states.

2. Materials and Methods

We conducted a retrospective cohort study using data from a U.S. integrated healthcare system (Northern California, Southern California, Northwest, and Hawaii regions) from 2010 to 2022. This healthcare system covers 12.6 million members throughout eight geographical regions in the US and has been shown to be demographically and socioeconomically representative [21,22,23]. This study was approved by the Kaiser Permanente institutional review board prior to commencement.

2.1. Inclusion and Exclusion Criteria

We used a combination of Current Procedural Terminology (CPT) codes from surgical logs and hospital billing data with medical device logs for breast implants and expanders from our integrated Electronic Health Record (EHR) to identify patients undergoing primary mastectomy with reconstruction. This information was used in combination to define the following mutually exclusive reconstruction categories: single stage: expander (no additional reconstruction), direct-to-implant, autologous, and two-stage reconstruction: expander–flap and expander–implant. We included only female patients aged eighteen and above and those undergoing surgery because of a diagnosis of breast cancer (ICD-9-CM: 174.0, 175, 196, 198.81, 233; ICD-10-CM: C44.5, C50, C77, C79, D05, V10, V86, V87). Patients without discharge status and those not discharged home were excluded. The final study sample included 23,301 primary mastectomy (Figure 1) cases with reconstruction procedures performed by 2639 surgeon teams at 38 healthcare centers.

2.2. Exposure of Interest

The exposure of interest was timing of breast reconstruction following primary mastectomy. IBR occurred within 0 to 14 days post-mastectomy, while delayed reconstruction, occurring at least 42 days post-mastectomy, was modeled as the reference group. We excluded reconstructions performed between 15 and 42 days post-mastectomy. Patients in this period would have a chance to heal their mastectomy skin flaps, reducing the risks compared to the immediate group, but they lacked the extended healing time of the delayed group [2]. Any cases where laterality could not be determined were excluded. Delayed reconstruction was modeled as the reference group.

2.3. Outcome of Interest

The primary outcome of interest was (1) time to first return to surgery of the breast for any reason and (2) multiple returns to surgery (1–5) of the breast for any reason after initial (first or second stage) reconstruction was complete. Patients with bilateral same-day surgery were counted as a single case and their reconstruction follow-up was calculated as time to first reconstruction for either breast or for assessment of multiple surgeries, only the same breast was used for standardization of comparison.

Secondary outcomes of interest were (1) time to first return to surgery of the breast for any reason and (2) multiple returns to surgery (1–5) of the breast stratified by reconstruction technique: single-stage (expander without reconstruction, autologous, and direct-to-implant) and two-stage approach (expander–flap and expander–implant). After surgery, patients were continuously monitored for death, outcome of interest, loss to follow-up through healthcare membership termination, or study end date (31 December 2022).

Procedures for patients undergoing autologous fat grafting (Current Procedural Terminology [CPT]: 15771, 15772) after primary reconstruction were not counted as a reoperation (n = 2322). These codes indicate procedures for grafting of autologous fat harvested by the liposuction technique to trunk, breasts, scalp, arms, and/or legs with a volume of 50 cc or less (CPT: 15771) and each additional 50 cc of harvested fat (CPT: 15772). Some surgeons routinely performed fat grafting after implant-based or autologous reconstructions, while others did not. This exclusion was based on the challenge of distinguishing between planned and unplanned fat grafting, and we opted to exclude procedures solely involving fat grafting from our outcome measures.

2.4. Radiotherapy

All administrative codes for radiation therapy were extracted for patients in the cohort that included the code description of breast and order dates were used in place of treatment date to account for procedures that occurred in outside or specialized freestanding facilities. Since patients could receive radiation therapy within and outside our healthcare system, and the administration of any radiation therapy required internal orders from our clinicians, we utilized the dates of order placement to initiate radiation therapy.

2.5. Covariates of Interest

The EHR was used to identify the various covariates at the time of index surgery, including age (continuous), race/ethnicity (Non-Hispanic White, Black, Hispanic, Asian, Missing/Other); body mass index (BMI, <18.5, 18.5–24.4, 25.0–29.4, 30.0–39.4, ≥39.5, and unknown); smoking (never, yes, quit, and other/missing); hospital region (Hawaii, Northern California, Northwest, and Southern California), bilateral same-day procedure (yes vs. no), year of mastectomy (2010–2022), American Society of Anesthesia (ASA) classification (1–2, 3–5, and missing), length of stay (days, continuous), radiotherapy (yes/no), and the following Elixhauser patient preoperative comorbid conditions in the year preceding surgery: total (0 vs. 1, 2, 3, 4, ≥5) and type (Yes vs. No) for congestive heart failure (CHF), diabetes mellitus, alcohol abuse, anemia, electrolyte and fluid disorder, peptic ulcer disease, peripheral vascular disease, pulmonary circulation disease, psychosis, renal insufficiency, and valvular disease. All covariates, except for the year of mastectomy, were assessed at the time of first reconstruction.

Reconstructions were categorized into a one-stage approach (direct-to-implant, flap with implant, primarily autologous reconstruction without an expander), two-stage approach (tissue expander followed by either implant or autologous reconstruction), and tissue expander with or without further reconstruction.

We used the standardized mean difference (SMD) to indicate covariate imbalance between reconstruction timing groups; covariates with a SMD > 0.20 were included in multivariable models.

2.6. Statistical Analysis

Descriptive statistics and incidence of first return to surgery were calculated. After the initial reconstruction, each patient experienced at various times a varying number of returns to the operating room representing up to five occurrences after the initial reconstruction until 7 July 2022. Crude incidence rates and 95% confidence intervals (CIs) were calculated as one minus the Kaplan–Meier estimator for first reoperation at 1-, 3-, 5-, and 10-year intervals. We used the Anderson–Gill Multiplicative Cox Proportional Hazards Model to account for inter-subject dependability and competing risk of mortality and loss to follow-up. The proportional hazard assumption was assessed by incorporating time as a multiplier for covariates; the proportional hazard assumption was met. Data analysis was performed using SAS Enterprise Guide (EG) 7.13 software (SAS Institute, Cary, NC, USA). All tests were two-sided, and statistical significance was defined as p < 0.05. Missing data were managed as a separate category for categorical variables.

3. Results

3.1. Overall

There were 23,301 total mastectomies with reconstruction procedures included in the analysis. Most patients were Non-Hispanic White, never smoked, had a BMI kg/m² of 18.5–24.4, and had the procedure in Southern California. The median patient age was 52, average follow-up time was 5.9 years (±3.8), and average time to first surgery was 1.0 years (±1.8).

In total, 78.4% (n = 18,276) of the cohort underwent immediate breast reconstruction following mastectomy and IBR increased over time from 61.7% of reconstruction procedures in 2010 to 94.3% of procedures in 2022. Compared to delayed patients, more IBR patients had never smoked and had bilateral procedures, procedures with expanders, and two-stage expander-to-implant reconstruction (

Table 1. Patient and surgical characteristics ¹ of women diagnosed with breast cancer undergoing primary mastectomy and reconstruction type at a Kaiser Permanente Facility between 2010 and 2022, where n = 23,301.

	Immediate Breast Reconstruction		Delayed Breast Reconstruction		SMD 2
	n = 18,276	78.4%	n = 5025	21.6%
Patient Characteristics
Age, Percentiles (25th 50th 75th)	43, 50, 59		45, 53, 61		−0.1889
Race/Ethnicity
Non-Hispanic White	10,236	56.0%	2855	56.8%
Black	1485	8.1%	372	7.4%
Hispanic	3518	19.3%	1165	23.2%
Asian	2780	15.2%	565	11.2%
Missing/Other	257	1.4%	68	1.4%
BMI, in kg/m2 (category)					−0.2560
<18.5	202	1.1%	43	0.9%
18.5–24.4	7114	38.9%	1404	27.9%
25.0–29.4	6279	34.4%	1814	36.1%
30.0–39.4	4380	24.0%	1632	32.5%
≥39.5	298	1.6%	124	2.4%
Unknown	3	0.0%	8	0.2%
Smoking					0.2522
Never	13,452	73.6%	3393	67.5%
Yes	217	1.2%	39	0.8%
Quit	4524	24.8%	1569	31.2%
Other/Missing	83	0.4%	24	0.5%
Hospital Characteristics					0.1219
Hospital Region
Hawaii	516	2.8%	256	5.1%
Northern California	7727	42.3%	1971	39.2%
Northwest	1188	6.5%	231	4.6%
Southern California	8845	48.4%	2567	51.1%
Bilateral Same-Day Surgery (yes)	12,512	68.5%	2393	47.6%	0.4162
Year of Mastectomy 3					0.3944
2010	830	61.7%	515	38.3%
2011	997	65.6%	524	34.5%
2012	1104	69.7%	479	30.3%
2013	1302	72.9%	484	27.1%
2014	1353	75.0%	450	25.0%
2015	1586	78.7%	430	21.3%
2016	1495	78.2%	417	21.8%
2017	1540	80.7%	369	19.3%
2018	1691	81.9%	375	18.2%
2019	1663	83.3%	334	16.7%
2020	1418	82.2%	308	17.8%
2021	1799	87.8%	250	12.2%
2022	1498	94.3%	90	5.7%
ASA Classification					−0.1524
1–2	15,347	84.2%	4092	81.4%
3–5	2682	14.5%	838	16.7%
Missing	247	1.3%	95	1.9%
Length of Stay (days) (mean, std)	1.3, 1.5		1.2, 2.1		0.0150
Radiotherapy (yes)	6425	35.2%	1881	37.4%	−0.0455
Reconstruction Type 4
Expander with or without Additional Reconstruction (yes)	13,204	72.3%	2561	51.0%	0.4485
Expander with no Additional Reconstruction (yes)	1198	6.6%	199	4.0%	0.1165
Single-Stage Reconstruction Procedure 4
Flap (yes)	2179	11.9%	1520	30.3%	−0.4610
Direct-to-Implant Reconstruction (yes)	2893	15.8%	944	18.8%	−0.0782
Flap plus Implant (yes)	2168	11.9%	641	12.8%	−0.0272
Two-Stage Reconstruction
Expander-to-Flap Reconstruction (yes)	1004	5.5%	148	3.0%	0.1270
Expander-to-Implant Reconstruction (yes)	11,002	60.2%	2214	44.1%	0.3274

SMD = standardized mean difference; ASA = American Society of Anesthesiologists; BMI = body mass index. ¹ Covariates are reported at the time of 1st reconstruction except where indicated. ² SMD of >0.20 was used to indicate covariate imbalance between immediate and delayed reconstruction groups for inclusion into multivariable models. ³ Row percentages. ⁴ Reconstruction-type covariates were assessed separately by reconstruction timing.

). Additionally, the percentage of IBR increased between 2010 and 2022 from 61.7% of procedures to 38.3% of procedures (SMD = 0.3944). No other covariates met the criteria for imbalance (SMD > 0.20) by timing and inclusion in multivariable models. A total of n = 1171 patients died during follow-up (IBR: 4.7% and delayed 6.1%) and n = 3003 patients terminated their membership (IBR: 12.8% and delayed: 12.9%). A higher percentage of patients in the IBR group had no Elixhauser comorbidities at the time of surgery (IBR 32.8% vs. delayed 28.2%), although none of the Elixhauser comorbidities assessed met the definition for covariate imbalance between the two groups (SMD > 0.20) (

Table 2. Comorbidity index hospitalization data * for breast cancer primary mastectomy patients by reconstruction type between 2010 and 2020, where n = 22,532.

Elixhauser Comorbidities	Immediate Breast Reconstruction (IBR)		Delayed Reconstruction		SMD *
	n = 17,643	78.3%	n = 4889	21.7%
No. of comorbidities					0.082
0	5787	32.8%	1379	28.2%
1	5328	30.2%	1594	32.6%
2	3599	20.4%	1002	20.5%
3	1553	8.8%	533	10.9%
4	829	4.7%	249	5.1%
5 or more	547	3.1%	132	2.7%
Alcohol Abuse	194	1.1%	73	1.5%	0.040
Anemia	1535	8.7%	381	7.8%	0.038
Congestive Heart Failure	159	0.9%	15	0.3%	0.081
Depression	1606	9.1%	455	9.3%	0.010
Diabetes Mellitus	2329	13.2%	694	14.2%	0.047
Electrolyte/Fluid Disorder	582	3.3%	132	2.7%	0.036
Peptic Ulcer Disease	35	0.2%	0	0.0%	0.055
PVD	776	4.4%	156	3.2%	0.058
PCD	71	0.4%	10	0.2%	0.028
Psychosis	1006	5.7%	352	7.2%	0.071
Renal Insufficiency	582	3.3%	93	1.9%	0.090
Valvular Disease	300	1.7%	54	1.1%	0.047

SMD, standardized mean difference; PVD, peripheral vascular disease; PCD, pulmonary circulation disease. * Comorbidity index hospitalization data were not available for 3.3% of procedures. Standardized mean differences > 0.20 were used to indicate imbalance and included for covariate adjustment in multivariable models.

).

The mean time to first post-reconstruction surgery was shorter in IBR compared to delayed reconstruction (1.0 years vs. 1.1). Overall, the mean operating time was longer in patients that underwent immediate reconstruction (IBR: 231.1, standard deviation [STD] ± 137.1 vs. delayed: 131.0, STD ± 60.3), and this time was higher for immediate reconstruction patients that underwent at least one reoperation (mean 241.3, STD ± 130.1), but lower in those who delayed reconstruction (mean: 126.0, STD ± 57.7). These differences potentially indicate the surgical complexity of patients who underwent immediate reconstruction. Patients who died during follow-up had a lower mean operating time regardless of reconstruction timing compared to the overall groups (IBR: 190, STD ± 137.5 and delayed: 129.3, STD ± 62.8).

3.2. Reoperation

The incidence of first reoperation was consistently higher in the IBR group, which increased over time, although the overall crude cumulative incidence of first return was higher in patients who delayed reconstruction compared to women who underwent immediate reconstruction (33.04%, 95% CI = 32.36–33.73 vs. 31.72%, 95% CI = 30.44–33.03) (

Table 3. Return to care and operating room (OR) outcomes of patients undergoing primary mastectomy procedures post-cancer diagnosis by reconstruction type at a Kaiser Permanente Facility Between 2010 and 2022, where n = 23,301.

	Breast Reconstruction Type		Adjusted HR (95% CI)
	Immediate	Delayed (Ref)
1st Reoperation	Cumulative Incidence (95% CI)
Overall %, (95% CI)	33.04% (32.36–33.73)	31.72% (30.44–33.03)	1.18 (1.12–1.25) 1
1 year	21.29% (20.69–21.90)	21.89% (20.77–23.07)
3 years	29.27% (28.58–29.97)	28.30% (27.06–29.58)
5 years	31.67% (30.95–32.41)	29.89% (28.61–31.21)
10 years	36.37% (35.48–37.29)	33.15% (31.72–34.61)
All Reoperations 3			1.00 (0.95–1.04) 2,3

¹ Cause-specific model accounting for death, membership termination and adjusted for BMI, smoking, bilateral procedures, year of mastectomy, expanders, flaps and expander-to-implant reconstruction. ² Anderson–Gills Multiplicative Hazards Model for multiple failure outcomes, accounting for patient death, membership termination, and adjusted for BMI, smoking, bilateral, year of mastectomy, expanders, flaps and expander-to-implant reconstruction. ³ Only the first 5 reoperations were captured. Bold indicates significance p < 0.05.

). After accounting for BMI, smoking status, year of mastectomy, bilateral procedures, any expander use, flaps, expander-to-implant reconstruction, and the competing risks of death and membership termination, patients who underwent IBR had an 18% higher cause-specific risk of first reoperation (HR = 1.18, 95% CI = 1.12–1.25), compared to those that delayed reconstruction. Figure 2 shows the post-operative probability of the first reoperation in women with IBR compared to those who delayed reoperation after accounting for BMI, smoking status, operating year, bilateral procedures, and implant characteristics as well as the competing risk of death and membership termination (Grey’s test: p < 0.0001).

There was no difference in risk of reoperation by timing (p > 0.05) when assessing up to the first five returns to OR (HR = 1.00, 95% CI = 0.95–1.04) (Table 3).

3.3. Stratified Analysis by Reconstruction Type

Analyses were stratified by reconstruction technique (incomplete, single-stage and two-stage) for first reoperation and up to five returns to OR in IBR patients with incomplete reconstruction. The risk for these IBR patients was 82% higher for the first return (HR = 1.82, 95% CI = 1.43–2.34) and 69% higher for multiple returns (HR = 1.69, 95% CI = 1.36–2.10). Single-stage reconstruction was also higher for IBR patients for first returns in autologous (HR = 1.19, 95% CI = 1.07–1.31) and direct-to-implant (HR = 1.47, 95% CI = 1.27–1.71) patients as well as multiple returns for the same reconstruction categories (autologous: HR = 1.09, 95% CI = 1.07–1.18 and direct-to-implant HR = 1.66, 95% CI = 1.49–1.86) (

Table 4. Implant, surgical, and pathological characteristics for first return to the operating room (OR) in patients undergoing primary mastectomy procedures post-cancer diagnosis by reconstruction type at a Kaiser Permanente Facility between 2010 and 2022 for 23,301 cases.

	Breast Reconstruction Type
	Immediate			Delayed (Ref)
	n=	Reoperations		n=	Reoperations		1st Reoperation	All Reoperations
		1st (%)	Mean (±SD)		1st (%)	Mean (±SD)	HR 1 (95% CI)	HR 2,3 (95% CI)
Incomplete Reconstruction with Expander
Expander	1198	652 (52.42)	0.80 ± 0.91	199	70 (35.18)	0.47 ± 0.77	1.82 (1.43–2.34)	1.69 (1.36–2.10)
Single-Stage Reconstruction
Autologous	2179	995 (45.66)	0.81 ± 0.93	1520	639 (42.02)	0.82 ± 0.99	1.19 (1.07–1.31)	1.09 (1.07–1.18)
Direct-to-Implant	2893	823 (28.45)	0.60 ± 0.94	944	228 (24.15)	0.44 ± 0.77	1.47 (1.27–1.71)	1.66 (1.49–1.86)
Two-Stage Reconstruction
Expander–Flap	1004	439 (43.73)	0.20 ± 0.49	148	67 (45.27)	0.15 ± 0.47	0.94 (0.72–1.23)	0.85 (0.69–11.04)
Expander–Implant	11,002	3130 (28.45)	0.10 ± 0.40	2214	590 (26.65)	0.07 ± 0.32	1.12 (1.02–1.23)	1.02 (0.99–1.17)

¹ Cause-specific model accounting for death, membership termination and adjusted for BMI, smoking, bilateral procedures, and year of mastectomy. ² Anderson–Gills Multiplicative Hazards Model for multiple failure outcomes, accounting for patient death, membership termination, and adjusted for BMI, smoking, bilateral procedures, and year of mastectomy. ³ Only the first 5 reoperations were captured. Bold indicates significance p < 0.05.

). Finally, two-stage reconstruction was only significantly (p < 0.05) higher for the first return to surgery in IBR patients that underwent expander–implant reconstruction (HR = 1.12, 95% 1.02–1.23) and was no different in IBR patients with expander–flap reconstruction for first returns (HR = 0.94, 95% CI = 0.72–1.23) or any of the two-stage reconstructions for multiple returns (expander–flap: HR = 0.85, 95% CI = 0.69–11.04 and expander–implant: HR = 1.02, 95% CI = 0.99–1.17) (Table 4).

4. Discussion

Despite incredible surgical and technical advancements, breast reconstruction remains a multistage process. The 10-year cumulative incidence of at least one reoperation in a year was 36.4% in IBR and 33.2% in delayed reconstruction patients. This finding reflects the complexity of breast reconstruction, and the technical difficulty compared with most other surgical specialties [24]. Knowledge of how the risk of reoperation is affected by reconstruction technique will help assist in surgical planning and patient information.

We found that IBR patients had an increased risk of first reoperation, while the risk of returning two to five reoperations was not significantly different by timing. Some studies of the impact of reconstruction timing on reoperation have found no difference between reconstruction timing [14], and others have found an increase in reoperations in patients who delayed reconstruction [17]. While Eon et al. found no statistical difference between immediate and delayed reconstruction (2.34 vs. 2.46, p = 0.690), a much smaller number of patients was included and competing risks as well as follow-up time were not incorporated into the analysis [14]. Losken et al. found delayed reconstructions required a higher number of secondary procedures in unilateral (p < 0.001) and bilateral (p = 0.03) reconstructions, although this study was conducted with data between 1975 and 2000 so updated reconstruction methodologies and recommendations that are currently used may be vastly different [17].

Rather than delineate reasons for reoperation, all reasons for return that involved breast reconstruction were included to give a more accurate representation of the first post-reconstructive landscape. Many studies have found lower risks of complications and failure in delayed compared to immediate reconstruction [3,25,26,27,28,29]. Additionally, more recent studies with a longer follow-up have found that delayed reconstruction patients have similar or superior patient satisfaction in psychosocial, sexual, or physical well-being realms [3,8,30,31,32,33]. These higher satisfaction scores in patients who delayed reconstruction may translate to a lower number of post-reconstruction surgeries as was seen in our study.

Planned post-mastectomy radiotherapy has negative effects on local tissues and damages healthy cells and as such is a major deciding factor in timing of reconstruction [34]. This can impair wound healing in irradiated tissues leading to a higher rate of implant loss and can negatively impact autologous tissue flaps [35]. We did not report a difference in radiation status by timing. This finding may reflect newer technology and conservative treatment choices [36,37,38], or that we were unable to assess radiotherapy dose and exact treatment days. Autologous reconstruction may be more impacted by radiotherapy and our method of measurement may be less precise compared to other studies that have found differences in this reconstruction technique category for delayed and immediate reconstruction [15,16].

Plastic surgeons have noticed a concurrent rise in patient-driven requests for bilateral mastectomy with IBR [39]. Contralateral breast surgery is performed more frequently in patients who undergo autologous reconstruction due to higher BMI, which necessitates contralateral mastopexy or breast reduction to restore symmetry when the flap used for the reconstructed breast does not match the contralateral breast [40,41]. We found an increased risk of reoperation for patients undergoing single-stage autologous reconstruction. Reconstruction involving contralateral breast surgery is much more demanding because it is associated with a significantly higher number of required surgical procedures, regardless of the reconstruction technique [15,17].

Breast reconstruction can be performed using several techniques including an expander/implant and/or autologous tissues. Autologous reconstruction has been shown to produce higher reconstruction patient satisfaction compared to implant-based reconstruction [42,43]. We did find that IBR patients had a higher risk of first reoperation and multiple returns to OR that underwent autologous based reconstruction.

A major disadvantage of implants is that they may look and feel less natural to the patient [44]. Many recent studies, including Ticha et al. have found implant-based reconstruction to require more surgeries (p = 0.012) and more time (p < 001) to complete reconstruction compared to no implant use [15]. This reflects our findings in IBR patients who underwent single-stage implant reconstruction compared to those who delayed reconstruction. Additional drawbacks to implant reconstruction involve the procedures used to remove capsular contracture that necessitate revision procedures such as capsulotomy, implant reposition, or implant exchange [41]. Our data show that the risk of reoperation for implant patients, including expander–implant and direct-to-implant procedures, was higher in IBR compared to delayed reconstruction. An explanation for this finding is that the more frequent need for implant revision may contribute to the lower satisfaction of patients with the outcome of implant reconstruction. Future breast implant technology improvements may increase patient satisfaction with implant reconstruction.

Strengths and Limitations

At baseline, in our study population, patient characteristics that may increase shorter term reoperation because of complications from wound healing, including advanced age, higher ASA or comorbidities, diabetes, and radiotherapy, were balanced between reconstruction timing groups (SMD < 0.20). This may explain some of the differences we found in our population compared to the literature regarding the risk of reoperation, particularly in the first year. Additionally, none of the Elixhauser comorbidities assessed were different between the reconstruction timing groups. Finally, unlike many other analyses that have assessed complications or reoperation-based outcomes, we accounted for competing events, including mortality and loss to follow-up. In the United States, our integrated healthcare system is unique in this ability to capture outcomes that may bias results due to incomplete information and are able to provide results based on longer-term follow-up. Additionally, we have detailed information on patient covariates and comorbidities that can reduce confounding in effect estimates. Mastectomy and reconstruction procedures were performed by many diverse surgical teams at 38 different medical centers so our findings may be more generalizable.

There are some noted limitations of our study. We did not assess differences in simple compared to more radical procedures or information on nipple reconstruction, which both have been known to affect the risk of return to the OR. Additionally, we did not have chemotherapy treatment information, particularly Tamoxifen, which may affect wound healing and increase patient morbidity and reoperation because of complications. We were unable to assess the reason for reoperation and identify reoperations for complications (e.g., infection, hematoma) as it would have required extensive chart review and was beyond the scope of our project. We were also unable to obtain information on the AJCC Stage at mastectomy and the timing and dosage of patient radiotherapy. Our analysis may have residual confounding. We cannot completely account for the complex factors and decision-making process between patients and clinical teams for breast reconstruction or reoperation.

5. Conclusions

In summary, this study is one of the largest assessments of reconstruction timing on the risk of reoperation including not only the first return but multiple returns to surgery (due to the complex nature of breast reconstruction processes) over a prolonged period of follow-up with detailed analysis accounting for radiotherapy and reconstruction techniques. Numerous studies in surgical areas substantiate the notion that preoperative expectations play a significant role in patient assessment of results and are strongly predictive of satisfaction and quality of life [32]. Understanding the differences in risk of reoperation among different reconstruction options is critical to making informed decisions about the type of breast reconstruction to undergo. This knowledge may also help patients work through complex emotional responses after reconstruction. A better understanding of these factors in our patients can help guide decision-making as IBR has become increasingly popular.