The 5-point Modified Frailty Index in Plastic and Reconstructive Surgery: Meta-analysis

Abstract

Background:

Frailty is regarded as a key predictor of adverse surgical outcomes. The 5-item modified frailty index (mFI-5) offers a simplified version of the validated 11-item risk stratification tool. Despite its use across surgical fields, its utility in plastic and reconstructive surgery remains underexplored. The purpose of this study was to evaluate and quantify the predictive value of frailty, as measured by the mFI-5 on postoperative outcomes in adult patients undergoing plastic and reconstructive surgery.

Methods:

PubMed, Embase, Web of Science, and Cochrane Library were systematically queried for studies comparing adverse events in frail (mFI-5 ≥ 2) versus nonfrail patients undergoing plastic and reconstructive surgery. Two independent reviewers performed study selection, data extraction, and risk of bias assessment using the risk of bias in non-randomized studies–of interventions tool, with evidence quality evaluated using the grading of recommendations, assessment, development and evaluation approach. Outcomes included surgical, medical, and any complications, readmission, reoperation, and mortality. Subgroup analyses were conducted by subject of study, database source, and study sample size.

Results:

Twenty-nine studies encompassing 302,641 patients were included. Frailty was significantly associated with increased odds of complications, readmission, reoperation, and mortality. Subgroup analyses by procedure type, database, and sample size consistently confirmed elevated risk across all outcomes, with varying degrees of statistical significance.

Conclusions:

Frailty, as measured by the mFI-5, is associated with an increased risk of postoperative complications, hospital readmission, reoperation, and mortality in plastic and reconstructive surgery patients. These findings support the integration of the index in clinical practice and preoperative patient evaluation and decision-making.

Takeaways

Question: Can the 5-item modified frailty index (mFI-5) effectively predict adverse postoperative outcomes in plastic and reconstructive surgery?

Findings: This systematic review and meta-analysis comprised 29 studies and more than 300,000 patients. Frailty, defined as mFI-5 greater than or equal to 2, was significantly associated with higher odds of surgical and medical complications, readmission, reoperation, and mortality. Subgroup analyses by procedure type and data source confirmed the predictive value across specialties.

Meaning: The mFI-5 is a simple and robust tool for identifying patients at risk for adverse postoperative outcomes in plastic surgery and should be clinically incorporated.

INTRODUCTION

Frailty is a recognized global health burden that is gaining international attention with the aging of the general population. Frailty is characterized by increased vulnerability to stressors and a global decline across numerous physiological systems.^1–3 It is associated with increased mortality, hospitalization, and admission to long-term intensive care.^1,2

Frailty has been suggested to be a valuable alternative to commonly used single-variable risk estimators in medicine and surgery. The hypothesis behind the suggestion is that aggregation of common risk factors can better estimate a patient’s susceptibility to subpar outcomes than focusing on individual factors.^4–6

Previously, age has been regarded as a critical factor in predicting complications. However, various reports have found no association between age and negative outcomes.^7,8

Plastic and reconstructive surgery is a broad surgical specialty encompassing a wide variety of surgical procedures and a heterogeneous patient population. Establishing a common risk estimator to identify patients at risk for complications and negative outcomes can aid in improving patient selection and ensuring safety.

The modified frailty index (mFI) was first introduced in 2013 by Velanovich et al.⁹ The authors correlated items from the Canadian Study of Health and Aging frailty index with preoperative clinical variables recorded by National Surgical Quality Improvement Program (NSQIP). Eleven variables were matched and constituted the original mFI in its 11-variable format, which was later significantly correlated with 30-day morbidity and mortality.

Over the years, with updates and modifications of the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database, certain variables have been removed, and as of 2015, only 5 of the original 11 variables from the mFI have remained: diabetes mellitus, hypertension requiring medication, chronic obstructive pulmonary disease, congestive heart failure within 30 days before surgery, and functional health status before surgery. To overcome this limitation and to maintain the updated relevance of the mFI, Subramaniam et al¹⁰ compared the predictive ability of a 5-factor mFI to the original 11-factor mFI for mortality, postoperative infection, and unplanned 30-day readmission. The authors concluded that both indexes are equally effective predictors in all surgical subspecialties, and the shortened version can adequately replace its predecessor. Despite its growing adoption, the utility of the 5-item mFI (mFI-5) in plastic and reconstructive surgery remains underexplored.

To date, no systematic reviews or meta-analyses have been performed on the mFI in plastic and reconstructive surgery. Previous reports tended to focus on individual procedures, such as breast or head and neck surgery, and lacked the comprehensive synthesis that is required to capture an overarching view of the subspecialty.^11,12 In this systematic review and meta-analysis, we aimed to evaluate and quantify the association between frailty, as measured by the mFI-5, and postoperative outcomes in adult patients undergoing plastic and reconstructive surgery. We further stratified the literature on the subject based on procedural subtype, sample size, and outcome.

MATERIALS AND METHODS

Ethical Considerations and Reporting Guidelines

This systematic review and meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and the Meta-Analysis of Observational Studies in Epidemiology Checklist (Fig. 1). (See appendix, Supplemental Digital Content 1, which displays the meta-analysis of observational studies in epidemiology checklist, https://links.lww.com/PRSGO/E641.)

Fig. 1.

PRISMA flowchart depicting the search and selection process. AUC, area under the curve.

Search Strategy and Selection Criteria

Four scientific databases were searched for eligible publications, including PubMed, Embase, Web of Science, and Cochrane Library databases on June 12, 2025.

The clinical question was structured using the population, intervention, control and outcome framework as follows:

• Population: Adult patients (≥18 y) undergoing plastic and reconstructive procedures.

• Intervention: Frail patients, defined as having an mFI-5 score of 2 or greater.

• Comparison: Nonfrail patients (mFI-5 < 2).

• Outcomes: Surgical complications, medical complications, any complications, readmission, reoperation, mortality, and length of hospital stay.

The complete search strategy and keyword adaptation for each database are provided in Supplemental Digital Content 2. (See appendix, Supplemental Digital Content 2, which displays the complete search strategy and keyword adaptation for each database, https://links.lww.com/PRSGO/E642.) Two independent reviewers (R.S. and Y.G.) individually screened titles and abstracts of all search results for eligible studies. The “snowballing” method, referring to the extraction of additional studies from the reference list of included publications, was used to increase outreach. Relevant publications were then screened based on their full text for inclusion. Inclusion criteria were studies comparing outcomes of plastic surgery procedures based on frailty status, including patient age 18 years or older, with frailty evaluated using the mFI-5, frailty defined as having an mFI score of 2 or greater and outcomes stratified by frailty status. Exclusion criteria included publication language other than English, no full text available, no mFI-based stratification, and frailty being defined as mFI scores of 1 or greater.

Articles were included in the study if both reviewers deemed them eligible. In instances of disagreements unresolvable by discussion, a third independent reviewer adjudicated the query (Yitzhak Skorochod). The complete screening process is shown in the PRISMA flowchart (Fig. 1).

Quality Assessment and Risk of Bias

Two reviewers (R.S. and Y.G.) independently evaluated the risk of bias of each study using the risk of bias in non-randomized studies–of interventions (ROBINS-I) tool. Quality of evidence was similarly evaluated using the grading of recommendations, assessment, development, and evaluation (GRADE) approach. Differences in scoring were resolved by discussion or by the inclusion of a third independent reviewer (Y.S.).

Data Extraction and Study Outcomes

From each included article, the following demographic information was extracted: first author, year of publication, country in which the study was performed, study design, database from which patients were recruited, study period, study design, subject of study, number of included patients, and proportion of frail patients (Table 1). Coprimary outcomes were the development of surgical complications, medical complications, or complications defined as “any complications.” Secondary outcomes were readmission, reoperation, and mortality (Table 2).

Table 1.

Summary of Variables Extracted From All Included Studies

Study ID, Author (Year)	Database	Period of Data	N Patients	Fragile Patients	Nonfragile Patients	Subject
Liu (2022)13	ACS-NSQIP	2013–2019	14,160	500	13,660	Breast
Jain (2021)14	ACS-NSQIP	2005–2018	11,852	514	11,338	Breast
Panayi (2022)15	ACS-NSQIP	2015–2019	40,415	1928	34,487	Breast
Lee (2020)16	ACS-NSQIP	2010–2015	575	421	154	Abdomen
Luo (2021)17	ACS-NSQIP	2012–2018	1254	364	890	General plastic surgery
Danko (2025)18	Emory	2015–2022	650	43	607	Breast
Hermiz (2021)19	ACS-NSQIP	2016–2018	22,700	1150	21,550	Breast
Hintze (2025)20	St. James Hospital, Ireland	2021–2023	159	40	119	Head and neck
Khan (2023)21	ACS-NSQIP	2010–2020	194	42	152	Lower extremity
Ash (2025)22	Emory	2002–2023	437	106	331	Abdomen
Hassan (2024)23	MD Anderson	2016–2022	1640	92	1548	Breast
Deldar (2024)24	Georgetown University	2011–2022	115	74	41	Lower extremity
Secanho (2025)25	Hospital das Clínicas da Faculdade de Medicina de Botucatu	2013–2020	229	71	158	General plastic surgery
Panayi (2024)26	ACS-NSQIP	2008–2021	34,571	2742	31,829	General plastic surgery
Panayi (2021)27	ACS-NSQIP	2015–2019	3795	598	3197	Head and neck
Knoedler (2025)28	ACS-NSQIP	2008–2021	96,586	12,092	83,550	Breast
Magno-Pardon (2022)29	ACS-NSQIP	2012–2018	10,550	469	10,081	Breast
Luo (2024)30	ACS-NSQIP	2014–2022	773	127	646	Head and neck
Akhavan (2024)31	ACS-NSQIP	2013–2019	421	14	407	Abdomen
Martinez (2022)32	ACS-NSQIP	2012–2016	5894	327	5567	General plastic surgery
Green (2025)33	Merative Marketscan Research Databases	2017–2022	16,287	1428	14,859	Breast
Brown (2024)34	Emory	1998–2020	547	55	492	Breast
Orgun (2025)35	ACS-NSQIP	2017–2022	5573	1022	4551	Head and neck
Othman (2025)36	New-Hyde Park	2017–2020	214	78	136	Head and neck
Jia (2024)37	ACS-NSQIP	2010–2020	5196	1466	3730	Lower extremity
Gonzalez (2025)38	ACS-NSQIP	2012–2020	219	20	199	Lower extremity
Huffman (2023)39	Georgetown University	2011–2022	115	74	41	Lower extremity
Chow (2023)40	ACS-NSQIP	2005–2019	25,215	1262	23,953	Breast
Desai (2024)41	ACS-NSQIP	2011–2021	2305	495	1810	Upper extremity

Table 2.

Summary of Outcomes Extracted From All Included Studies

Study ID, Author (Year)	Any Complication, OR (95% CI)	Surgical Complication, OR (95% CI)	Medical Complication, OR (95% CI)	Readmission, OR (95% CI)	Reoperation, OR (95% CI)	Mortality, OR (95% CI)
Liu (2022)13	N/A	1.85 (1.26–2.73)	3.11 (1.66–5.84)	2.46 (1.48–4.07)	1.11 (0.58–2.10)	N/A
Jain (2021)14	N/A	1.67 (1.38–2.01)	N/A	1.66 (1.17–2.35)	1.30 (1.01–1.66)	N/A
Panayi (2022)15	N/A	1.44 (1.21–1.72)	2.26 (1.72–2.98)	1.68 (1.39–2.03)	1.45 (1.23–1.70)	0.95 (0.06–16.22)
Lee (2020)16	1.34 (1.09–2.15)	1.26 (1.08–2.20)	N/A	N/A	N/A	N/A
Luo (2021)17	1.33 (1.04–1.72)	N/A	N/A	N/A	N/A	N/A
Danko (2025)18	N/A	2.15 (1.15–4.03)	N/A	N/A	0.63 (0.08–4.81)	N/A
Hermiz (2021)19	N/A	1.45 (1.14–1.85)	2.11 (1.63–2.72)	1.72 (1.27–2.33)	1.46 (1.10–1.95)	N/A
Hintze (2025)20	N/A	1.63 (0.79–3.35)	2.32 (1.10–4.89)	2.08 (0.75–5.80)	2.66 (1.25–5.65)	3.13 (0.61–16.20)
Khan (2023)21	2.04 (1.01–4.10)	1.23 (0.44–3.41)	2.78 (1.36–5.67)	1.99 (0.69–5.77)	3.55 (1.47–8.54)	1
Ash (2025)22	N/A	1.28 (0.79–2.07)	N/A	N/A	N/A	N/A
Hassan (2024)23	N/A	3.11 (1.98–4.88)	3.31 (1.72–6.38)	2.91 (1.85–4.59)	2.04 (1.25–3.35)	N/A
Deldar (2024)24	N/A	1.63 (0.65–4.11)	N/A	N/A	N/A	N/A
Secanho (2025)25	N/A	N/A	N/A	10.02 (1.02– 98.61)	N/A	N/A
Panayi (2024)26	2.19 (2.03–2.37)	2.29 (2.11–2.49)	3.13 (2.79–3.51)	1.83 (1.58–2.13)	1.67 (1.51–1.85)	3.80 (2.63–5.50)
Panayi (2021)27	1.47 (1.23–1.75)	1.26 (1.05–1.51)	1.68 (1.35–2.09)	1.09 (0.81–1.46)	1.10 (0.88–1.37)	1.20 (0.62–2.33)
Knoedler (2025)28	1.31 (1.21–1.41)	1.63 (1.45–1.83)	2.14 (1.77–2.59)	2.70 (2.37–3.07)	0.74 (0.65–0.85)	3.08 (1.55–6.13)
Magno-Pardon (2022)29	1.80 (1.48–2.18)	1.59 (1.28–1.97)	1.88 (1.49–2.38)	N/A	N/A	N/A
Luo (2024)30	6.57 (4.34–10.03)	2.85 (1.24–6.16)	N/A	N/A	0.95 (0.22–2.91)	4.14 (0.51–27.1)
Akhavan (2024)31	7.43 (2.25–24.61)	1.34 (0.15–12.20)	N/A	2.36 (0.28–19.85)	58.3 (3.22–1052.06)	N/A
Martinez (2022)32	N/A	2.12 (1.51–2.97)	3.09 (2.44–3.91)	1.68 (1.15–2.46)	1.64 (1.27–2.16)	N/A
Green (2025)33	1.36 (1.20–1.55)	1.36 (1.19–1.56)	2.18 (1.65–2.87)	N/A	1.14 (1.02–1.28)	N/A
Brown (2024)34	N/A	1.87 (0.82–4.23)	N/A	N/A	N/A	N/A
Orgun (2025)35	N/A	1.37 (1.22–1.55)	N/A	1.30 (1.05–1.62)	1.36 (1.14–1.62)	2.06 (1.20–3.53)
Othman (2025)36	2.14 (1.17–3.91)	3.46 (1.93–6.22)	N/A	N/A	2.03 (1.04–3.99)	N/A
Jia (2024)37	1.02 (0.89–1.15)	1.18 (1–1.39)	3.00 (2.64–3.41)	2.56 (1.30–5.03)	N/A	N/A
Gonzalez (2025)38	4.1 (1.6–10.54)	2.68 (0.80–8.91)	N/A	2.75 (0.7–10.7)	4.82 (1.72–13.48)	N/A
Huffman (2023)39	N/A	1.19 (0.77–1.84)	N/A	N/A	N/A	N/A
Chow (2023)40	1.62 (1.42–1.84)	N/A	N/A	N/A	N/A	N/A
Desai (2024)41	2.71 (2.19–2.36)	1.44 (0.66–3.13)	N/A	1.14 (0.61–2.14)	1.48 (0.94–2.35)	N/A

N/A, not available.

Statistical Analysis

Continuous variables were reported as mean values with the corresponding SD. Categorical variables were reported as frequencies and percentages.

Odds ratios (ORs) were calculated for each study when not directly reported, using the prevalence of the outcome in each group. The natural logarithm of each OR (log [OR]) and its corresponding standard error were computed to facilitate meta-analytic pooling. Pooled ORs were calculated using a random-effects model, weighting each study by the inverse of the variance of the log-transformed OR.

Heterogeneity between studies was assessed using the I² statistic and τ². Publication bias was evaluated based on funnel plots and Egger test. Subgroup analyses were performed based on the databases used, sample size, and subject of study.

RESULTS

Baseline Variables

Overall, 29 studies were included in the meta-analysis presented in this study,^13–41 resulting in 302,641 included patients, of whom 9.1% were classified as frail based on the 5-mFI. Most studies analyzed data from the ACS-NSQIP and were performed by first authors based in the United States. Breast surgery was the focus area in 11 studies, followed by head and neck surgery and lower extremity in 5 studies.

Overall Analyses

Pooled OR for any postoperative complication was 2.71 (95% confidence interval [CI] 2.35–3.13), with an I² value of 64%. For medical complications, the pooled OR was 2.64 (95% CI, 2.31–3.01, I² = 0%). Surgical complications were associated with an OR of 1.65 (95% CI, 1.49–1.83; I² = 0%). OR for reoperation was 1.29 (95% CI, 1.11–1.49; I² = 0%). Readmission was significantly increased in the frail group, with OR of 1.91 (95% CI, 1.64–2.23; I² = 0%). For mortality, the pooled OR was 2.71 (95% CI, 1.78–4.13, I² = 0%) (Fig. 2).

Fig. 2.

Forest plots of pooled OR for adverse events in the overall cohort. A, Any complications. B, Surgical complications. C, Medical complications. D, Readmission. E, Reoperation. F, Mortality. IV, inverse variance; SE, standard error.

Subgroup Analysis

Subgroup analysis of studies investigating frailty in breast surgery was further conducted. The OR for any complication was 1.41 (95% CI, 1.16–1.72; I² = 0%). The OR for medical complications was 2.15 (95% CI, 1.92–2.40; I² = 0%), and for surgical complications, 1.56 (95% CI, 1.41–1.73; I² = 0%). Readmission was also increased, with an OR of 2.23 (95% CI, 1.71–2.91; I² = 0%). In terms of reoperation, an OR of 1.10 (95% CI, 0.87–1.40) was observed, and for mortality, the pooled OR was 2.88 (95% CI, 0.09–91.75; I² = 0%). (See figure, Supplemental Digital Content 3, which displays forest plots of pooled OR for adverse events in the breast surgery cohort. A, Any complications. B, Surgical complications. C, Medical complications. D, Readmission. E, Reoperation. F, Mortality, https://links.lww.com/PRSGO/E643.)

Subgroup analysis of studies investigating frailty in head and neck surgery revealed that the OR for surgical complications was 1.39 (95% CI, 1.08–1.80; I² = 0%). Reoperation was also increased, although not in a statistically significant manner, with an OR of 1.30 (95% CI, 1.00–1.69; I² = 0%). The OR for any complication was 2.22 (95% CI, 0.33–14.83), with low heterogeneity (I² = 30%). Medical complications were associated with an OR of 1.72 (95% CI, 0.57–5.21; I² = 0%), readmission with an OR of 1.24 (95% CI, 0.86–1.79; I² = 0%), and mortality with an OR of 1.79 (95% CI, 0.96–3.33; I² = 0%). (See figure, Supplemental Digital Content 4, which displays forest plots of pooled OR for adverse events in the head and neck surgery cohort. A, Any complications. B, Surgical complications. C, Medical complications. D, Readmission. E, Reoperation. F, Mortality, https://links.lww.com/PRSGO/E644.)

In studies focusing on lower extremity surgery, the OR for medical complications was 3.07 (95% CI, 2.66–3.55; I² = 0%). Surgical complications had a pooled OR of 1.21 (95% CI, 1.06–1.39; I² = 0%). The pooled OR for any complication was 1.82 (95% CI, 0.74–4.48), with high heterogeneity (I² = 79%). Reoperation had an OR of 2.04 (95% CI, 0.48–8.67; I² = 0%), and readmission had an OR of 1.78 (95% CI, 0.88–3.60; I² = 0%). (See figure, Supplemental Digital Content 5, which displays forest plots of pooled OR for adverse events in the lower extremity surgery cohort. A, Any complications. B, Surgical complications. C, Medical complications. D, Readmission. E, Reoperation, https://links.lww.com/PRSGO/E645.)

For studies querying the ACS-NSQIP database, the OR for any complication was 1.75 (95% CI, 1.37–2.24; I² = 67%). The OR for medical complications was 2.66 (95% CI, 2.27–3.11; I² = 0%), and for surgical complications, 1.68 (95% CI, 1.48–1.90; I² = 0%). The pooled OR for reoperation was 1.31 (95% CI, 1.08–1.59; I² = 0%), and for readmission, 1.89 (95% CI, 1.60–2.24; I² = 0%). The OR for mortality was 2.70 (95% CI, 1.65–4.40; I² = 0%). (See figure, Supplemental Digital Content 6, which displays forest plots of pooled OR for adverse events in the ACS-NSQIP cohort. A, Any complications. B, Surgical complications. C, Medical complications. D, Readmission. E, Reoperation. F, Mortality, https://links.lww.com/PRSGO/E646.)

In a subgroup of studies with sample sizes greater than 1000 patients, which included 9–14 studies per outcome, the OR for any complication was 1.64 (95% CI, 1.28–2.10; I² = 78%). For medical complications, the OR was 2.64 (95% CI, 2.28–3.05; I² = 0%), and for surgical complications, 1.66 (95% CI, 1.44–1.91; I² = 0%). The pooled OR for reoperation was 1.27 (95% CI, 1.06–1.52; I² = 0%), and for readmission, 1.91 (95% CI, 1.58–2.31; I² = 0%). Mortality was significantly elevated as well, with an OR of 2.68 (95% CI, 1.48–4.84; I² = 0%). (See figure, Supplemental Digital Content 7, which displays forest plots of pooled ORs for adverse events in cohorts with sample size greater than 1000 patients. A, Any complications. B, Surgical complications. C, Medical complications. D, Readmission. E, Reoperation. F, Mortality, https://links.lww.com/PRSGO/E647.)

Risk of Bias Assessment

Of the 31 included studies, 14 were rated as having a serious overall risk of bias, 16 as moderate, and 1 as low. A visualization of the risk of bias analysis is presented in Figure 3.

Fig. 3.

ROBINS-I risk of bias assessment of included studies.

Quality of Evidence Assessment

Certainty of evidence was assessed using the GRADE framework and ranged from low to very low across all included outcomes. Of the total assessments, 17 studies (55%) were rated as having very low certainty and 14 studies (45%) as low certainty.

DISCUSSION

With the growing complexity of plastic and reconstructive surgical procedures and the aging population, identifying adequate predictors of postoperative adverse events is of utmost importance. In this systematic review and meta-analysis, we sought to evaluate the predictive value of the mFI-5 in predicting postoperative adverse events of plastic and reconstructive surgical procedures. Integration of data from 29 studies reflecting 300,000 patients demonstrated that the mFI-5 is significantly associated with an increased risk for any complications, surgical or medical, readmission, reoperation, and mortality across all plastic surgery procedures. Individual studies previously reported frailty and the mFI-5 to be useful predictors of complications in specific plastic surgery procedures; here, we present the first study to synthesize these findings into a unified pooled estimate across the specialty. The results and derived conclusions from our analyses correspond with previous works in other medical fields.

Khan et al⁴² implemented the mFI-5 in patients undergoing parastomal hernia repair with records documented in the ACS-NSQIP between 2015 and 2020. The authors found that frail patients had a higher risk of renal, cardiovascular, pulmonary, and hematologic complications, as well as readmissions, prolonged hospitalizations, discharge to an assisted care facility, and overall mortality.

Similar results were reported by Chiarella et al⁴³ when analyzing patients undergoing mastectomy procedures without subsequent reconstruction. Higher rates of overall surgical complications, pneumonia, deep vein thrombosis, and acute kidney failure were found to be statistically associated with higher frailty status.

Subgroup analyses demonstrated variability in the magnitude of effect based on procedure type and data source. In breast surgery, frailty was associated with an increased risk for all outcomes, with no heterogeneity between studies, suggesting high internal consistency. Among lower extremity studies, similar results were observed. However, high heterogeneity and lack of statistical significance were observed in the analysis of “any complications,” likely signifying differences in outcome definitions between studies. The overall high pooled OR for medical complications and readmission can be a derivative of the complexity and comorbidity of this patient population. Analysis of head and neck studies demonstrated similar results with a higher frequency of clinical but not statistically significant outcome measures, potentially due to smaller sample sizes and patient-specific confounders such as malnutrition and prior chemoradiation therapy.

As we hypothesized, the heterogeneity in certain subgroup analyses stems from variability in patient populations, procedural complexity, and outcome definitions. Random effect models were used to mitigate the influence as well as additional sensitivity analyses to explore the contributors.

Analyses stratified by studies reporting ACS-NSQIP analyses demonstrated significance results, highlighting the robustness of conclusions. High heterogeneity was again observed for the outcome of “any complications,” likely the result of differences in outcome definitions between the various studies.

Findings from this study support the integration of frailty screening using the mFI-5 in the preoperative setting of patients undergoing plastic and reconstructive surgery. The tool offers a simple, fast, and validated approach to adequately identifying patients at high risk for adverse events and detrimental complications from the surgical procedures. The predictive power across diverse subspecialties further supports its validity despite heterogeneous patient characteristics. The incorporation of the mFI-5 into preoperative consultations can highlight patients who can benefit from alternative treatment modalities or preoperative optimization of comorbidity control. Although our findings suggest that the mFI-5 is a useful predictor of adverse outcomes, these findings should be interpreted cautiously given the low certainty of underlying evidence.

Current literature is filled with numerous risk estimators and predictors that aim to identify patients at risk for postprocedural adverse events. The most frequently used among them is the Charlson Comorbidity Index, aimed at predicting 10-year mortality by considering 17 comorbid conditions. Contrary to the simplicity and quick assessment of the mFI-5, the Charlson Comorbidity Index focuses on chronic disease burden and requires thorough medical evaluation and assessment.⁴⁴

Key strengths of our study include the large pooled sample size, rigorous adherence to reporting guidelines (PRISMA), risk of bias assessment (ROBINS-I), and transparent quantification of evidence quality using validated tools (GRADE). Notably, this meta-analysis is novel, and its findings are generalizable to many plastic surgery procedures and practices.

Nonetheless, no study comes without limitations. In our study, we are compelled to mention the low quality of evidence observed in included studies, namely due to their observational nature and high risk of bias. The observational nature of studies may also result in publication bias stemming from the exclusion of small sample studies describing neutral effects.

Additionally, complications may have varied in their definition or standard of recording and could have impacted the quality of analysis. The former is a typical inherent limitation of large registry studies. Retrospective registry analysis allows for an overview of a large subset of patients from multiple institutions, which can be very appealing. However, reliance on registries can result in coding errors, missing data, and nondifferential misclassification biases. Despite their disadvantages, analysis of these large databases provides valuable insight, and the pros and cons must be weighed for each research question.

It is also worth mentioning that a large portion of studies included in this review had a very large sample size. Although the former is considered beneficial and increases the power of conclusions, it could also lead to statistical significance of minuscule differences in outcome prevalence.

Another important limitation that should be disclosed is the geographic concentration of included studies, with the vast majority being based in the United States and focusing on ACS-NSQIP data. This restricts the external validity of our findings to global or resource-limited plastic surgery settings.

As stated, all studies included in these analyses are retrospective in nature. Future directions should include the prospective inclusion of the mFI-5 in preoperative consultations and evaluation of its potential to identify patients at risk and those requiring more thorough preoperative management.

In conclusion, frailty as measured by the mFI-5 is associated with increased risk of postoperative complications, hospital readmission, reoperation and mortality in plastic and reconstructive surgery procedures. These findings support the incorporation of the mFI-5 in clinical practice and preoperative patient evaluation and decision-making. Future studies must compare the predictive ability of the mFI-5 to other commonly used comorbidity indices and determine which tool best highlights patients at greater surgical risk.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.

ACKNOWLEDGMENT

The authors thank Dr. Yitzhak Skorochod for his assistance with data curation and for resolving disagreements during article screening and quality-of-evidence assessment.