Frailty, commonly defined as a deficit in physiologic reserve with decreased physical function beyond what is expected with normal aging, is one metric used to evaluate a patient’s fitness for surgery [1,2,3,4]. Frailty measurements have long been a part of the medical literature and more recently have gained attention as a predictive metric for postoperative outcomes [1, 5,6,7]. The tools used to measure frailty vary in complexity, ranging from physical tests (e.g., timed get up and go) to cumulative deficit models, such as the modified frailty index (mFI) which measures frailty based on a number of comorbidities [8, 9]. The mFI has been shown to have prognostic value across multiple surgical specialties [9,10,11,12,13,14].

There are a few studies that specifically examine frailty and its effect on outcomes after ventral hernia repair (VHR), one of the most common procedures performed in the USA. Previous studies examining frailty in the VHR population report poor outcomes in frail patients, but studies to date have been unable to control for hernia-specific variables known to effect outcomes in VHR [15,16,17,18,19]. For instance, frail patients in these studies may have more complex hernias, making it difficult to discern if their poor outcomes were a result of their physical state or the nature of their hernia. This is a critical gap in the current literature, given that complication rates are recognized to vary significantly by hernia size [17, 19, 20]. Examining the association of frailty with operative outcomes in the context of predictive factors, such as hernia size, is an important step to improve perioperative risk assessment and counseling for this common general surgery procedure.

We sought to evaluate the association of preoperative frailty with short-term outcomes in VHR using the 5-factor modified frailty index, a cumulative deficit frailty model demonstrated to correlate with operative outcomes, while taking into consideration patient- and hernia-specific factors. We used data from a population level, clinically nuanced hernia data registry, the Michigan Surgery Quality Collaborative Hernia Registry (MSQC-HR) that captures patient- and hernia-specific surgical data in the state of Michigan. We hypothesized that patients with higher frailty scores would have worse short-term postoperative outcomes independent of hernia size and characteristics.

Methods

This secondary analysis of deidentified data was deemed exempt from review by the Institutional Review Board of the University of Michigan. The requirement for informed consent was waived because of a lack of identifying information. This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines [21].

Data source and patient population

Data in this study were obtained from the Michigan Surgical Quality Collaborative Hernia Registry (MSQC-HR), a statewide surgical quality improvement collaborative of 70 hospitals in Michigan [22,23,24]. The MSQC is funded by Blue Cross Blue Shield of Michigan and maintains a clinical registry of prospectively collected patient characteristics, perioperative and intraoperative processes of care, hernia characteristics, and 30-day clinical outcomes. The MSQC-HR data reported in this manuscript include all payors (Medicare, Medicaid, Private, Uninsured). The process by which the MSQC abstracts data, ensures validity and reliability, and samples cases has been previously described [24].

Included patients were adults (age ≥ 18 years) that underwent ventral hernia repair surgery collected by the MSQC between January 1, 2020 and January 31, 2022. Operations included were open, laparoscopic, and robotic ventral and incisional hernia repairs. Procedures were identified using Current Procedural Terminology codes: 49,560, 49,561, 49,565, 49,566, 49,570, 49,572, 49,585, 49,587, 49,590, 49,652, 49,653, 49,654, 49,655, 49,656, and 49,657. Patients were excluded if they had an incomplete record for patient demographics, comorbidities, or hernia-specific data (Supplemental Fig. 1).

Outcomes and covariates

The primary outcomes of this study were any 30-day postoperative complications. Secondary outcomes included serious 30-day postoperative complications, surgical site infection (SSI), non-home discharge, and a composite of post-discharge adverse events (emergency department visit, readmission, or reoperation). Serious complications were defined using a previously published definition of serious complications [25]. A full list of any and serious 30-day complications can be found in Table 1 [25, 26].

Table 1 Complications description
Full size table

The covariate of interest was patient frailty, which was defined using the 5-factor modified frailty index (mFI5) scoring system. The mFI5 has been previously validated across multiple surgical specialties, including the VHR population, to correlate with patient outcomes [14, 15]. The variables included in the 5-factor modified frailty index include (1) congestive heart failure within 30 days prior to surgery, (2) the presence of non-insulin-dependent or insulin-dependent diabetes mellitus, (3) a history of chronic obstructive pulmonary disease (COPD) or pneumonia within 30 days prior to surgery, (4) partially dependent or totally dependent functional health status, and (5) hypertension requiring medication. Each factor counts as one point in the mFI5 score. The mFI5 scores were stratified into three categories: no frailty (mFI5 = 0), moderate frailty (mFI5 = 1), and severe frailty (mFI5 ≥ 2). This stratification of mFI5 categories has been previously utilized and validated in the surgical literature [10, 14, 27, 28]. This condensed 5-item modified frailty index is relevant to the VHR population as the included variables are well-known risk factors for complications [17, 29, 30].

Patient demographic data included age, sex, payor status, and race. Clinical characteristics included body mass index (BMI), American Society of Anesthesiologists (ASA) classification, and comorbidities, including diabetes, smoking, obstructive sleep apnea (OSA), COPD, functional status, cancer, and hypertension. Hernia-specific characteristics included previous hernia repair, hernia width, hernia location, wound classification as defined by the Centers for Disease Control (CDC) guidelines, mesh use, mesh type, and mesh fixation [31, 32]. Other operative characteristics included component separation use and type, surgical priority (elective versus urgent/emergent), and surgical approach (open versus minimally invasive). Our secondary outcome was SSI.

Statistical analysis

Descriptive analysis was used to describe characteristics of the cohort, surgical, and other hernia-specific variables. Univariate analysis was performed using the Chi-squared and analysis of variance (ANOVA) tests where appropriate. Multivariable analyses were performed to evaluate factors associated with the primary and secondary outcomes, including the factors mFI5 score, age, sex, race, BMI, ASA class (I–II, III–IV), smoking, cancer, OSA, previous hernia repair, hernia width, hernia location, surgical priority, and surgical approach (open versus minimally invasive). Age, BMI, and hernia width were used as continuous variables in regression analysis. The individual variables included in the mFI5 score were not separately included in the multivariable regression. Race was not included in the surgical site infection regression as it was a predictor of failure for the model. For the multivariable regression analysis, the hernia location variables suprapubic and infraumbilical were combined.

All analysis were performed using StataSE version 16.1 (StataCorp, Inc., College Station, TX). All tests of statistical significance were 2-sided with an alpha of 0.05. Data were analyzed in February and March 2022.

Results

Patient characteristics

A total of 4406 patients that underwent VHR were included in this analysis. Baseline patient demographics, comorbidity, and hernia characteristic data are listed in Table 2. The mean (SD) age of the cohort was 55 (15) years, 2015 (46%) were female, and 3591 (82%) white. The mean (SD) BMI of the cohort was 33 (8) kg/m2. Most patients had an ASA class of I–II (54%). The median hernia size was 2.5 cm (1.5–4.0 cm). The most common comorbidity was hypertension, followed by OSA and smoking. Most patients were undergoing primary, elective VHR (83%). A majority of VHR were classified as clean (94%) and repaired with an open approach (58%) and synthetic mesh (83%). Operative characteristics of the cohort are described in Table 3.

Table 2 Patient and hernia characteristics
Full size table
Table 3 Operative characteristics
Full size table

A total of 2077 (47%) patients had no frailty (mFI5 = 0), 1604 (36%) were moderately frail (mFI5 = 1), and 725 (17%) were severely frail (mFI5 ≥ 2). The most common comorbidity contributing to mFI5 scores was hypertension, followed by diabetes. Patients with an mFI5 score ≥ 2 were on average older, had higher BMI, had higher surgical risk (83% ASA classification III–IV), and had larger hernias.

Complications

The unadjusted incidence of primary and secondary outcomes increased in a stepwise fashion as frailty increased (Table 4). Postoperative complications occurred in 141 (3%) of patients and serious complications occurred in 52 (1%). Compared to patients with no frailty, patients with severe frailty had a higher incidence of complications (7 versus 2%, p < 0.001), serious complications (3 versus 1%, p < 0.001), SSI (3 versus 1%, p < 0.001), non-home discharge (9 versus 2%, p < 0.001), and post-discharge adverse events (16 versus 8%, p < 0.001). Overall, the most common complications were SSI, postoperative transfusion, and postoperative sepsis. In patients with severe frailty, the most common complications were SSI, pneumonia, and postoperative transfusion.

Table 4 Unadjusted complication rates
Full size table

On multivariable regression analysis, severe frailty was independently associated with increased odds of any complication (aOR 3.12, 95% CI 1.78–5.47), serious complication (aOR 5.25, 95% CI 2.17–13.19), SSI (aOR 3.41, 95% CI 1.58–7.34), and post-discharge adverse events (aOR 1.70, 95% CI 1.24–2.33), compared to patients with no frailty. This translates to predicted probabilities of 5.11% (95% CI 3.55–6.68) for any complication, 2.64% (95% CI 1.39–3.89) for serious complications, 2.74% (1.51–3.95) for surgical site infection, and 13.78% (11.18–16.38) for post-discharge adverse events (Fig. 1). Severe frailty was not associated with non-home discharge (aOR 1.64, 95% CI 0.93–2.90). Independent of hernia size, compared to patients with patients with no frailty, patients with severe frailty had a higher predicted probability of any complication (2-cm hernia: 5.34% (95% CI 6.90–3.78) versus 1.26% (95% CI 1.71–0.80); 5-cm hernia 6.65% (95% CI 8.42–4.87) versus 1.60% (95% CI 2.14–1.05); Fig. 2), and serious complications (2-cm hernia: 2.64% (95% CI 3.79–1.49) versus 0.44% (95% CI 0.71–0.16); 5-cm hernia 3.09% (95% CI 4.31–1.86) versus 0.52% (95% CI 0.82–0.21).

Fig. 1
figure 1

Predicted probability of postoperative outcomes stratified by mFI5 score, Black = mFI5 = 0; Blue = mFI5 = 1; and Red = mFI5 = 2, *p-value < 0.05 compared to mFI5 = 0 (Color figure online)

Full size image
Fig. 2
figure 2

Predicted probability of any postoperative complication over hernia size, Black = mFI5 = 0; Blue = mFI5 = 1; and Red = mFI5 = 2 (Color figure online)

Full size image

To evaluate if surgical approach influenced the primary and secondary outcomes in patients with severe frailty, we performed a separate multivariable analysis in the population of severely frail patients. Within the severely frail population, an open surgical approach was associated with an increased odds of any complication (aOR 2.52, 95% CI 1.14–5.56) and surgical site infection (aOR 4.52, 95% CI 1.22–16.75). An open surgical approach was not associated with increased odds of a serious complication (OR 2.02, 95% CI 0.71–5.75).

Discussion

In this cohort of patients undergoing VHR, we found that frailty was associated 30-day complications, independent of hernia size. Hernia size has not been controlled for in previous studies of frailty in this population and therefore this analysis adds a level of nuance to the existing literature. These data are also consistent with previous reports describing an increased risk of postoperative complications in frail surgical patients [1, 6, 16, 33]. Severe frailty was also associated with experiencing a post-discharge adverse event. In short, frailty appears to be an important predictor of a surgical patient’s perioperative trajectory. These results suggest that inclusion of frailty assessments, regardless of hernia size, in the preoperative period may improve risk stratification, inform postoperative resource allocation, and importantly, improve patient counseling.

Frailty is a measure of a person’s physiologic reserve to withstand stressors [1, 2, 4]. An abundance of studies have demonstrated the association of frailty with worse postoperative complications and increased mortality [16, 27, 28, 33,34,35]. For example, in a large analysis of 71,455 patients undergoing ambulatory hernia repair collected in the American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database, increased frailty was associated with increased odds of experiencing postoperative complications [16]. As a result, frailty assessments have been suggested for inclusion in the preoperative assessment of surgical patients. The current study demonstrates independent of hernia size that these same trends apply to the VHR population. Although a small hernia repair may not be technically complex, a frail patient’s postoperative course may be. Severely frail patients undergoing VHR are more likely to experience surgical complications and adverse post-discharge events. Additionally, independent of patient- and hernia-related factors, an open surgical approach was associated with an increased risk of postoperative complications in severely frail patients. Discussion of the increased risk of complications using an open surgical approach in the preoperative counseling visit may better inform a frailer patient’s decision-making for VHR.

Prior work has demonstrated an association between frailty and post-discharge adverse events, such as non-home disposition, reoperation, and hospital readmission [36,37,38]. Our results support these findings and suggest that frailty may be a predictor of both surgical outcomes and post-discharge events in the VHR population. In this study, the most common complications in severely frail patients were SSI, pneumonia, and postoperative transfusion. Accordingly, patients may benefit from increased focus on pre- and postoperative interventions to reduce SSI, such as preoperative antiseptic bathing, glycemic control, and maintenance of perioperative normothermia [39]. Additionally, added focus on pulmonary hygiene in patients with severe frailty may be of added benefit [40].

Frailty assessments that utilize a cumulative comorbidity model, such as the mFI5 used in this study, are simpler compared to physical assessments of frailty and can be used to screen for higher-risk surgical patients more easily. Frailty screening and subsequent focus of resources toward frail patients can improve outcomes in surgical patients without negatively impacting non-frail patients [41]. In addition, presurgical optimization programs for hernia patients have been shown to improve management of higher-risk patients and increase operative yield in the hernia population [42,43,44]. For example, at our institution patients are screened in the preoperative period for frailty and other high-risk characteristics and, if positive, are referred to our hernia optimization clinic. In the optimization clinic, patient health characteristics, functional status, and frailty can be further assessed, and patients are subsequently provided guidance and resources to meet their health needs. This intervention has resulted in improved optimization of hernia repair patients and is an impactful example of a preoperative intervention for higher-risk patients [42].

The results of this study must be considered within the context of its limitations. First, the observational design of this study introduces the possibility of selection bias and does not allow for causal inferences from the reported data. This study analyzed a database of Michigan patients undergoing VHR limiting the generalizability of these results to the general population. Additionally, these data were collected during the COVID-19 pandemic, introducing selection bias for whom was given the option of VHR versus watchful waiting. Patients who underwent watchful waiting of their VHR are not accounted for in these data and this potentially introduces further bias. These data also do not report surgical site occurrences (e.g., seroma, dehiscence) and lack long-term follow-up data which are necessary to report hernia recurrence in this population. In addition, the number of patients with missing variables is high. Still, it is critical to understand the predictors of immediate complications given that up to 20% of patients undergoing VHR experience a complication [17]. These limitations are due to the recent implementation of the MSQC-HR data collection and long-term follow-up and reporting of this data is planned. Additionally, we only report clinical outcomes in this study and not patient-reported outcomes which are important to consider in hernia repair. Finally, the frailty score used in this study relies upon variables that reflect comorbidities of the patient. This score does not consider the physical exam or other potential frailty measures that account for physical activity such as the Timed Up and Go Test or hand-strength that have been shown to be predictive of postoperative outcomes [5, 6]. However, the mFI5 has been validated in the general surgery population and is a strong metric to analyze surgical patients and outcomes in large, population databases [14, 15, 27].

Conclusion

In this cohort of patients undergoing VHR, when controlling for patient-, operative-, and hernia-specific factors, increased frailty was associated with a worsening risk of postoperative complications across all hernia sizes. In the severely frail population, independent of patient- and hernia-related factors, an open surgical approach was associated with an increased risk of postoperative complications. These findings highlight the importance of patient frailty assessments in the preoperative setting. Addressing frailty in the perioperative period may improve patient counseling and outcomes for this common general surgery procedure.