Abstract
Background:
The meniscus has a proven role in load transmission, stability, and prevention of osteoarthritis (OA). Research suggests that meniscal repair may improve long-term joint health as compared with meniscectomy, but higher reoperation and readmission rates may be incurred.
Purpose:
To examine 90-day readmissions, reoperations, and health care utilization for symptomatic OA for patients who underwent meniscal repair vs meniscectomy.
Study Design:
Cohort study; Level of evidence, 2.
Methods:
Patients undergoing 1 of 2 index procedures, meniscal repair or meniscectomy, were identified in the Premier Health Database. Three endpoints were investigated: 90-day readmission, reoperations, and health care utilization for symptomatic OA. The demographics for each index procedure group were matched by propensity score matching. Chi-square tests of independence were used to compare endpoints between the index procedure groups, and subanalyses were performed for patients aged <40 vs ≥40 years.
Results:
After matching, 36,386 patients (36,386 knees; 18,193 with meniscectomy and meniscal repair each) were included in this study (mean ± SD age, 38 ± 17 years [meniscectomy group, 38.45 ± 17.29; repair group, 38.10 ± 17.37]; mean Charleston Comorbidity Index, 0.12 ± 0.38 [meniscectomy group, 0.12 ± 0.38; repair group, 0.12 ± 0.37]). The index procedures had similarly low rates of 90-day readmission (1.1% for repair and 1.9% for meniscectomy). Patients undergoing repair were significantly less likely to necessitate reoperation, whether the index procedure or another type (2.3% vs 4.3% for meniscectomy; P < .0001). Patients undergoing meniscectomy were significantly more likely to receive OA treatment within 2 years of meniscectomy as opposed to meniscal repair (9.2% vs 6.7%; P < .0001), and this trend was maintained for both age groups.
Conclusion:
This study demonstrated that the index procedures had similarly low rates of 90-day readmission. Patients undergoing repair were significantly less likely to necessitate any reoperation. OA treatment was significantly more likely in patients who had meniscectomy as opposed to meniscal repair, and this trend was maintained for both age groups. These data may support consideration for meniscal preservation in all patients, regardless of age; however, future studies are needed to further stratify age-appropriate recommendations and tear patterns.
In recent years, the number of meniscal repairs performed annually in the United States has increased.2,50 Utilizing the American Board of Orthopaedic Surgery database, Wasserburger et al 50 found that meniscal repair volume significantly increased after 2010, and the number of surgeons performing meniscectomies significantly decreased. This surgical trend may be attributed to continued improvements in meniscal repair technologies and a better understanding of the benefits of preserving the meniscus.6,8
The meniscus functions in load distribution and stability.18,51 The menisci transmit >50% of the force acting on the knee, lessening the burden on articular cartilage.1,7,18,39,48,49 They also act as secondary stabilizers of the knee joint, preventing anterior-posterior and rotational laxity.4,18,20,38,48 As appropriate knee biomechanics depend on meniscal function, meniscal loss can transmit excess force to the articular cartilage and lead to progression of osteoarthritis (OA).14,26,30,44
As the meniscus’ role in load transmission, stability, and prevention of OA has been better understood, the benefits and risks of meniscectomy and meniscal repair have been investigated. Initial investigations suggested that meniscectomy has lower reoperation rates, and the 2 procedures have little difference in long-term outcomes.10,29,35,42,52 However, as the meniscal repair technique has evolved, many studies have shown lower rates of knee OA and reoperation as compared with meniscectomy.25,27,29,41,43,46,52 In a 2023 meta-analysis, Migliorini et al 27 found that meniscal repair significantly lowered the rate of advanced knee OA and progression to total knee arthroplasty (TKA). In 2011, Paxton et al 29 reported that meniscal repairs had higher reoperation rates than meniscectomy but better long-term outcomes. Yet, in one of the few database studies on the topic, Sochacki et al 41 revealed significantly lower reoperation rates in patients with meniscal repair when compared with those with meniscectomies in 2020. Sochacki et al contributed to a growing body of evidence suggesting that meniscal repair is beneficial for patients aged ≥40 years, evolving previous orthopaedic dogma.15,29,37,41,47
The purpose of this study was to analyze meniscal repair vs meniscectomy using a large, nationally representative database. Information was collected as related to 90-day readmissions, reoperations, and health care utilization for symptomatic OA. We hypothesized that meniscal repairs may incur more reoperations of the same type as the index procedure but will result in better long-term joint outcomes as compared with meniscectomy, specifically reducing health care utilization related to OA and the need for reoperation of the index procedure or another type. Furthermore, we hypothesized that there is no difference in the 3 endpoints for patients aged <40 vs ≥40 years.
Methods
Data Collection
The Premier Healthcare Database (PHD) was searched from January 1, 2017, to March 31, 2023. Each year, the PHD collects health care utilization data from a geographically and socioeconomically diverse pool of >1200 US hospitals. 31 In the deidentified database, each patient is assigned a unique identifier so that the same patient can be tracked across multiple visits.
Data were included if patients underwent 1 of 2 index procedures: outpatient meniscal repair or meniscectomy (see Appendix 1 for Current Procedural Terminology codes). Patients were excluded if they underwent bilateral procedures or the operative leg could not be identified via a Current Procedural Terminology code. Patients were also excluded if they underwent multiple procedures during the index therapy. Patient demographics and procedure-related variables were recorded, including age at baseline, sex, race, marital status, and comorbidities at baseline (eg, body mass index >30 kg/m2, peripheral vascular disease, and diabetes).
Three endpoints were investigated: readmission at 90 days, reoperation rate (index procedure vs other reoperation), and health care utilization for symptomatic OA within 2 years of the index procedure. Reoperations were categorized as a patient undergoing a subsequent meniscectomy or meniscal repair after the index procedure vs other procedures after the index procedure, such as meniscal transplant or arthroplasty (Appendix 1). Readmissions were assessed at 90 days. In addition to OA diagnosis codes, indications of symptomatic OA included physical therapy (defined as physical therapy that continued 6 months past the index therapy), intra-articular hyaluronic acid injections, and intra-articular steroid injections. While these metrics are not surrogates for OA, they represent health care utilization for persistent symptoms, a labor-intensive need for continued treatment, and health care utilization after an index procedure.
Statistical Analysis
Descriptive analyses of patient demographics, comorbidities, and procedure-related variables were generated for the meniscal repair and meniscectomy cohorts. To control for selection bias between the groups, the demographics for a cohort of meniscal repair cases were matched to those of meniscectomy cases by propensity score matching. Matching variables included hospital-specific data such as size, teaching status, region, urban/rural setting, and ratio of cost to charges. Patients were also matched according to race, year of procedure, and Charleson Comorbidity Index. To examine the effectiveness of propensity score matching in balancing baseline characteristics within the matched cohorts, standardized differences were calculated for the matched variables with a greedy algorithm under 1:1.
Chi-square tests of independence were used to compare our 3 endpoints—readmission at 90 days, reoperation rate, and health care utilization for symptomatic OA—between the index procedure groups. Further comparisons were made for patients aged <40 vs ≥40 years. All statistical analyses were performed in SAS Version 9.4 (SAS Institute). Statistical significance was defined a priori as P < .05.
Results
Patient Demographics
After matching, 36,386 patients were studied: 18,193 underwent meniscectomy and 18,193 underwent meniscal repair. Before propensity matching, the mean age of our patient population was 51.5 years (SD, 15.5) in the meniscectomy group and 37.5 years (SD, 17.5) in the meniscal repair group (Table 1). The sample was overwhelmingly White (79.6% and 74.9% in the meniscectomy and meniscal repair groups, respectively) and skewed male (51.7% and 53.2%). After matching, patient demographics were appropriately similar between the groups, with a mean age of approximately 38 years (SD, 17) and mean Charleson Comorbidity Index score of 0.12 (SD, 0.38) (data for all matched variables can be found in Table 1 and Appendix 2).
Patient Demographics a
Data are presented as mean ± SD or No. (%). See Appendix 1 for other matched variables. CCI, Charlson Comorbidity Index.
Patient Outcomes
The 90-day readmission rates for the meniscectomy and meniscal repair groups were low at 0.87% and 0.63%, respectively (P = .0074; Table 2). Patients undergoing repair were significantly less likely to necessitate reoperation of the index procedure type (2.3% vs 4.3% for meniscectomy; P < .0001). This relationship remained significant for patients <40 years old (2.8% vs 4.4%; P < .0001) and ≥40 years old (1.9% vs 4.2%; P < .0001). Patients undergoing repair were also significantly less likely to necessitate reoperation of the index procedure or another type, including TKA (1.1% vs 1.9%; P < .0001). Again, this relationship remained significant for patients aged <40 years (0.08% vs 0.25%; P = .0046) and ≥40 years (2.3% vs 3.7%; P < .0001).
Meniscectomy vs Meniscal Repair Outcomes a
Data are presented as No. (%). Bold indicates P < .05. ICD-10-CM, International Classification of Diseases, Tenth Revision, Clinical Modification; OA, osteoarthritis.
Hyaluronic acid injections, physical therapy, steroid, or ICD-10-CM code. Per protocol, health care utilization related to OA required a 2-year follow-up period; therefore, propensity matching was performed separately.
The No. value for OA matching is noted for each OA data point.
Patients undergoing meniscectomy were significantly more likely to utilize health care for OA treatment within 2 years of meniscectomy as opposed to meniscal repair (9.2% vs 6.7%; P < .0001). This trend was maintained for patients aged <40 years (4.9% vs 3.1%; P < .0001) and ≥40 years (14.9% vs 11.4%; P < .0001).
Discussion
The major findings of our study demonstrated that patients undergoing meniscal repair were significantly less likely to necessitate any reoperation, whether of the index procedure or another type, than their counterparts undergoing meniscectomy. This trend was maintained for patients aged <40 and ≥40 years. OA treatment was significantly more likely in patients who had meniscectomy as opposed to meniscal repair (9.2% vs 6.7%; P < .0001), with the trend maintained for patients aged <40 and ≥40 years. Notably, patients undergoing meniscectomy are more likely to be treated for degenerative meniscal tears, often involving concomitant OA and other age-related joint changes; to account for this trend, patient demographics, including age, were matched. 13 These data support the notion that meniscal repair is beneficial in patients older than 40 years. Historically, literature has suggested that the vascularity of the meniscus diminishes at the age of 40 years, and meniscal repair may not be successful.11,21 Our data add to the growing literature supporting meniscal repair and preservation in patients older than 40 years.22,53 Given these findings, we can reject our null hypothesis that there is no difference in index procedure outcome for patients aged <40 vs ≥40 years.
The present study expands on foundational work assessing outcomes of meniscectomy as compared with meniscal repair. Paxton and colleagues’ hallmark systematic review 29 aggregated the data from 95 studies utilizing either intervention.3,10,35,36 This review found that the reoperation rate after meniscectomy was approximately 4% as opposed to 23% for meniscal repair; however, meniscal repair resulted in better Lysholm grades and fewer radiographic changes 10 years after surgery. Of note, Paxton et al did not examine specific reoperation procedures, only aggregate reoperation rates. Our study revealed similar trends with regard to joint health, noting decreased health care utilization for symptomatic OA within 2 years of surgery in patients undergoing meniscal repair. Our work challenges previous studies’ findings, specifically Paxton and colleagues’ systematic review, suggesting that reoperation rates are higher among patients with repair. 46 Patients undergoing repair were significantly less likely to necessitate reoperation (2.3% vs 4.3% for meniscectomy; P < .0001). This aligns with Sochacki and colleagues’ previous findings. 41 One potential explanation for this discrepancy is the era in which the data were collected. Paxton et al completed their systematic review in 2011, but as previously mentioned meniscal repair surgery increased after 2010 according to Wasserburger and colleagues’ American Board of Orthopaedic Surgery study. 50 With increasing attention to meniscal preservation, meniscal repair devices have also modernized with focus on better biomechanical strength as well as circumferential suturing patterns.5,33 Moreover, biological augmentation (ie, platelet rich plasma, bone marrow venting) has been increasingly employed and proven to aid in healing rates for meniscal repair surgery.12,16 This present study collected data from 2017 to 2023 and thus may provide a more accurate representation of patient-reported outcomes, reoperation rates, as well as 90-day readmission rates.
This study was primarily limited by the nature of the PHD. Our reoperation rates were lower than those published in previous literature, suggesting that data may have been lost.28,29 The PHD accounts only for data at approximately a quarter of US hospital admissions and cannot follow patients who seek further care at a nonparticipating institution. The lower-than-expected reoperation rates could also be a function of the limited follow-up of 2 years. The 2-year study period limits analyses of these index procedures and subsequent OA-related health care utilization. Furthermore, physical therapy, intra-articular hyaluronic acid injections, and intra-articular steroid injections were considered indications of symptomatic OA in addition to OA diagnosis codes. These metrics provide an estimate of health care utilization related to OA; however, findings would be more robust if supported by a narrative account of OA in the electronic medical record. Given the 2-year study period and OA endpoints representing midterm proxies for symptomatic OA, future research should obtain long-term endpoints such as conversion to TKA to make more definitive conclusions about long-term OA progression and structural changes. The PHD does not include other data points of interest, such as preexisting chondral damage, surgical technique, and meniscal tear type and laterality. Preexisting chondral damage was not available in the database; yet, given the equal distribution in patient characteristics (age, body mass index), we assume that chondral changes would be evenly balanced. The PHD does not record surgical techniques; therefore, we could not control for this. Moreover, the database does not provide detailed information on meniscal tear type or laterality. Some meniscal tear types are more amenable to repair and healing than other tear types. For example, tears in the peripheral zone are more conducive to meniscal repair owing to the zone's increased vascularity, and in terms of morphology, vertical longitudinal tears have some of the highest rates of successful repair.9,34,45 Still, research suggests that patients of similar body mass index, age, and other demographic factors are likely sustain meniscal tears in similar patterns; given this consideration and the appropriate matching of our cohort, we believe that our analysis provides a valid estimate of meniscal injury and repair success.19,23,25,32 Nevertheless, we advocate for future research that considers these parameters—including tear morphology, existing chondral damage, and surgical technique—in the investigation of meniscal repair vs meniscectomy in various patient populations. Despite these limitations, this study is strengthened by its large, nationally representative sample and attention to multiple postoperative metrics, including reoperation and health care utilization for symptomatic OA.
Given meniscal repair's potential to reduce health care utilization for symptomatic OA, future medical care and research must prioritize interventions to further improve meniscal repair. Specifically, the promise of repair bioaugmentation, including marrow venting and platelet-rich plasma, should be evaluated and implemented in practice as standard of care.12,16,40 The cost-effectiveness of meniscal repair as compared with partial meniscectomy has been described. Feeley et al 17 performed a decision-analytic Markov disease progression model looking at meniscal repair vs meniscectomy failure rates and subsequent development of OA and the need for TKA. This model found that payers (ie, insurance companies) could save approximately $43 million annually if 10% of current meniscectomies were performed as meniscal repairs. Lester et al25 reported similar benefits in the setting of an anterior cruciate ligament reconstruction; in their model, there was an estimated gain of 18.00 quality-adjusted life-years after anterior cruciate ligament reconstruction with meniscal repair as compared with 17.16 with partial meniscectomy (an increase of 0.84).
Conclusion
Our study demonstrated that the 2 index procedures had similarly low rates of 90-day readmission. Patients undergoing repair were significantly less likely to necessitate any reoperation. OA treatment was significantly more likely in patients who had meniscectomy as opposed to meniscal repair, and this trend was maintained for both age groups. These data may support consideration for meniscal preservation in all patients, regardless of age; however, future studies are needed to further stratify age-appropriate recommendations and tear patterns.
Footnotes
Appendix
Patient Demographics and Other Matched Variables
| Non–Propensity Score Matched | Propensity Score Matched | |||||
|---|---|---|---|---|---|---|
| Meniscectomy |
Meniscal Repair |
P Value | Meniscectomy |
Meniscal Repair |
P Value | |
| Hispanic | <.0001 | .0034 | ||||
| No | 288,013 (79.65) | 15,432 (76.74) | 14,193 (78.01) | 14,088 (77.44) | ||
| Unknown | 42,031 (11.62) | 2103 (10.46) | 1635 (8.99) | 1818 (9.99) | ||
| Yes | 31,544 (8.72) | 2574 (12.8) | 2365 (13) | 2287 (12.57) | ||
| Marital status | <.0001 | .097 | ||||
| Married | 200,490 (55.45) | 7675 (38.17) | 7092 (38.98) | 7195 (39.55) | ||
| Other | 23,832 (6.59) | 1355 (6.74) | 1322 (7.27) | 1211 (6.66) | ||
| Single | 135,451 (37.46) | 10,956 (54.48) | 9684 (53.23) | 9680 (53.21) | ||
| Unknown | 1815 (0.5) | 123 (0.61) | 95 (0.52) | 107 (0.59) | ||
| Hospital region | <.0001 | ≥.999 | ||||
| Midwest | 113,213 (31.31) | 5856 (29.12) | 5386 (29.6) | 5386 (29.6) | ||
| Northeast | 60,859 (16.83) | 2750 (13.68) | 2413 (13.26) | 2413 (13.26) | ||
| South | 145,855 (40.34) | 9043 (44.97) | 8361 (45.96) | 8361 (45.96) | ||
| West | 41,661 (11.52) | 2460 (12.23) | 2033 (11.17) | 2033 (11.17) | ||
| Teaching hospital | <.0001 | ≥.999 | ||||
| No | 221,728 (61.32) | 11,037 (54.89) | 10,089 (55.46) | 10,089 (55.46) | ||
| Yes | 139,860 (38.68) | 9072 (45.11) | 8104 (44.54) | 8104 (44.54) | ||
| Hospital setting | <.0001 | ≥.999 | ||||
| Rural | 53,226 (14.72) | 2241 (11.14) | 1869 (10.27) | 1869 (10.27) | ||
| Urban | 308,362 (85.28) | 17,868 (88.86) | 16,324 (89.73) | 16,324 (89.73) | ||
| No. of beds at hospital | <.0001 | ≥.999 | ||||
| 000-099 | 54,560 (15.09) | 2967 (14.75) | 2661 (14.63) | 2661 (14.63) | ||
| 100-199 | 74,880 (20.71) | 3613 (17.97) | 3275 (18) | 3275 (18) | ||
| 200-299 | 77,773 (21.51) | 4110 (20.44) | 3731 (20.51) | 3731 (20.51) | ||
| 300-399 | 55,321 (15.3) | 2787 (13.86) | 2493 (13.7) | 2493 (13.7) | ||
| 400-499 | 37,850 (10.47) | 1972 (9.81) | 1725 (9.48) | 1725 (9.48) | ||
| ≥500 | 61,204 (16.93) | 4660 (23.17) | 4308 (23.68) | 4308 (23.68) | ||
| Cost type | <.0001 | .5011 | ||||
| Procedural | 221,584 (61.28) | 12,974 (64.52) | 11,951 (65.69) | 11,890 (65.35) | ||
| Ratio of cost to charge | 140,004 (38.72) | 7135 (35.48) | 6242 (34.31) | 6303 (34.65) | ||
Final revision submitted October 8, 2025; accepted October 9, 2025.
One or more of the authors has declared the following potential conflict of interest or source of funding: R.M.P. is a consultant for Smith & Nephew. L.N. and S.D. are employees of Smith & Nephew. S.L.S. has received consulting fees from Smith & Nephew, Linvatec Corporation, Vericel Corporation, Flexion Therapeutics Inc, Ceterix Orthopaedics, Joint Restoration Foundation, Olympus America, Bioventus LLC, LifeNet Health, Kinamed Inc, Pacira Therapeutics Inc; royalties from Linvatec Corporation and ConMed Corporation; honoraria from Vericel Corporation, Flexion Therapeutics, and Joint Restoration Foundation; a grant from Linvatec Corporation; speaking and travel fees from Arthrex and Synthes GmbH; support for education from Elite Orthopedics and Evolution Surgical; and hospitality payments from Aesculap Biologics and Stryker Corporation. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
