Abstract
Introduction/background
Given the uncommon nature of the multi-ligament injured knee, a better understanding of its outcomes may be achieved by contemporary evaluation of a national administrative database. We aim to identify risk factors for 90-day major complications, 90-day readmissions, and subsequent knee surgery after a multi-ligament knee reconstruction.
Methods
The Humana Claims Database was queried for all patients undergoing a multi-ligament knee reconstruction procedure from 2007 to 2016 using International Classification of Diseases (ICD)-9 procedure codes and Current Procedural Terminology (CPT) codes. Outcomes of interest included Center of Medicaid and Medicare Services (CMS)- reportable 90-day complications, 90-day readmission and subsequent ligament reconstruction or total knee arthroplasty. Predictive factors studied included demographics (ie. age, sex, tobacco use) comorbidity burden and operative factors (ie. surgical setting, diagnosis of knee dislocation, concomitant meniscal repair or meniscectomy).
Results
588 multi-ligament knee reconstruction procedures were identified. The 90-day readmission rate 8.3%. The rate of subsequent ligament surgery was 7.1%. On multivariate regression analysis, a diagnosis of knee dislocation was associated with a four times greater likelihood of readmission and a two and a half greater likelihood of subsequent ligament surgery. Concomitant meniscectomy was associated with higher likelihood of subsequent total knee arthroplasty (9.1, 1.4–67.0) and outpatient setting of surgery compared to inpatient hospital setting was associated with reduced likelihood of readmission.
Conclusion
While uncommon, the patient population undergoing multi-ligament knee surgery is relatively diverse. Understanding common factors predisposing this population to adverse events following surgery may assist in improved outcomes.
Keywords: Knee dislocation, Multiligamentous knee injury, Knee ligament reconstruction, Quality improvement
1. Introduction
There are a number of unique circumstances regarding multi-ligament knee injuries that pose challenges to management and clinical investigation. First, the reported incidence of these injuries varies in the literature given that diagnosis can be missed due to the heterogeneous mechanisms that can produce this injury.1 The consensus in the literature is that the multi-ligament knee injuries are relatively uncommon. Recent epidemiological evidence demonstrating an incidence of 0.072 events per 100 patient-years and another study reporting that isolated anterior cruciate ligament (ACL) reconstruction are over 60 times as common as reconstructive procedures for multi-ligament knee injuries.1,2 Consequently, the existing literature draws largely from cases series of single institution experiences.3, 4, 5, 6
Second, patients with multi-ligament knee injuries are incredibly heterogenous in their presentation, severity of injury, and approach to surgical management. Significant neurovascular damage can be encountered in these patients requiring adjunctive surgical procedures.7 Ligament reconstruction may be staged based on the clinical scenario and surgeon preference, and the timing of surgery can be incredibly varied.8,9,18
The current state of the art on multi-ligament knee injuries has been advanced by the combined effort and experiences of multiple centers, but these individual studies are limited in their scope to capture the epidemiologic profile of these injuries.3, 4, 5,10, 11, 12 A better understanding of trends in presentation and management of these injuries, as well as their association with outcomes of interest, may be achieved by contemporary evaluation of a national administrative database. This study aims to identify risk factors for 90-day major complications, 90-day readmissions, and subsequent knee surgery after a multi-ligament knee reconstruction.
2. Methods
2.1. Patient selection
The Humana Claims Database contains medical, prescription, and laboratory claims data from 20.9 million privately-insured and Medicare Advantage patients from 2007 to 2016. The database was queried for all patients undergoing a multi-ligament knee reconstruction procedure using International Classification of Diseases (ICD)-9 procedure codes and Current Procedural Terminology (CPT) codes (Appendix 1). A multi-ligament procedure was defined by ICD-9-CM procedure codes 81.45 and 81.46 recorded with the same procedure date or CPT codes for both a cruciate ligament procedure and collateral ligament procedure with the same procedure date. Patients actively enrolled in the database at a minimum of 6 months prior to and 18 months after surgery were included. Patients under the age of 15 and over the age of 65 were excluded.
2.2. Patient factors
Basic demographic factors including age, sex, location of surgery (inpatient vs outpatient setting) were studied. Individual comorbidities, such as obesity, tobacco use, and diabetes, were queried and comorbidity burden was assessed with Charlson Comorbidity Index, a validated scoring system incorporating differentially weighted comorbidities and age classes. Additional factors studied included prior placement of external fixator, diagnosis of knee dislocation, concomitant neuroplasty, concomitant repair of peroneal nerve, concomitant meniscal repair, and concomitant meniscectomy.
2.2.1. Outcomes
4 main outcomes were studied. Outcomes occurring 90 days after surgery included Center for Medicare and Medicaid Services (CMS)-reportable complications (myocardial infarction, pneumonia, venous thromboembolism, sepsis, post-operative bleeding, wound infection, septic arthritis) and readmission. Subsequent ligament surgery was defined as the presence of a procedural code for ipsilateral cruciate or collateral ligament surgery 6 months after the index multi-ligament procedure, to account for deliberately staged ligament reconstructions. Subsequent total knee arthroplasty (TKA) was analyzed as well.
2.3. Statistical analysis
Factors of interest were included in logistic regression analysis to identify associations with the four main outcomes (90-day CMS complications, 90-day readmission, subsequent ligament surgery, subsequent TKA). Odds ratios and 95% confidence intervals were reported for significant covariates detected via logistic regression. Significance was defined as P-value less than .05.
3. Results
588 patients were identified undergoing a multi-ligament knee reconstruction. Their demographics are reported in Table 1. Females comprised 37.1% of the group. 20.1% of the patients were obese. 16.8% of the entire group underwent surgery in an inpatient setting. 179 patients (30.4%) had a diagnosis of knee dislocation. 15 patients had prior external fixation.
Table 1.
Patient Gender, Age and Associated Co-Morbidities.
| N | % Total | ||
|---|---|---|---|
| Female Gender | 218 | 37.1% | |
| Age | 15 to 19 | 144 | 24.5% |
| 20 to 24 | 66 | 11.2% | |
| 25 to 29 | 48 | 8.2% | |
| 30 to 34 | 51 | 8.7% | |
| 35 to 39 | 43 | 7.3% | |
| 40 to 44 | 54 | 9.2% | |
| 45 to 49 | 37 | 6.3% | |
| 50 to 54 | 35 | 6.0% | |
| 55 to 59 | 24 | 4.1% | |
| 60 to 64 | 16 | 2.7% | |
| Obesitya | 118 | 20.1% | |
| Tobacco Use | 131 | 22.3% | |
| Diabetes | 72 | 12.2% | |
| Prior Placement of External Fixation | 15 | 2.6% | |
| Diagnosis of Knee Dislocation | 179 | 30.4% | |
| Concomitant Neuroplasty | 31 | 5.3% | |
| Concomitant Repair of Peroneal Nerve | 12 | 2.0% | |
| Concomitant Meniscal Repair | 84 | 14.3% | |
| Concomitant Meniscectomy | 191 | 32.5% | |
| Setting of Surgery | Inpatient Hospital | 99 | 16.8% |
| Hospital-based Outpatient Department | 326 | 55.4% | |
| Ambulatory Surgical Center | 163 | 27.7% | |
| Charlson Comorbidity Index | Mean, standard deviation | .6 (1.35) |
CMS-reportable complication occurred within 90 days in 34 patients (5.8%). 90-day readmission rate was 8.3%. This study identified 42 patients (7.1%) with subsequent ligament surgery, and 11 patients (1.9%) with subsequent total knee arthroplasty (TKA). These findings are reported in Table 2.
Table 2.
CMS Complication and Readmission.
| Outcome | N | %Total | |
|---|---|---|---|
| 90-day CMS Complication | 34 | 5.8% | |
| 90-day Readmission | 49 | 8.3% | |
| Subsequent Knee Surgery | Ligament Reconstruction | 42 | 7.1% |
| Total Knee Arthroplasty | 11 | 1.9% |
Regression analysis identified factors associated with CMS complications, readmissions, subsequent ligament surgery, and subsequent TKA. Comorbidity burden, determined by Charlson Comorbidity Index, was associated with increased likelihood of CMS complications (1.4, 1.0–1.8). Factors associated with increased likelihood of readmission included obesity (4.2, 1.8–9.5) and knee dislocation diagnosis (4.0, 1.7–9.5). Surgery at a hospital-based outpatient department (HOPD) and ambulatory surgical center (ASC), relative to an inpatient setting, was associated with decreased likelihood of readmission. Knee dislocation (2.5, 1.2–5.2), prior placement of external fixator (1.3, 1.1–1.7), and concomitant neuroplasty (1.5, 1.37–1.9) was associated with increased likelihood of subsequent ligament surgery. Finally, concomitant meniscectomy (9.1, 1.4-67.-0) was associated with increased likelihood of subsequent TKA. These findings can be found in Table 3.
Table 3.
Factors Associated with Complcations, Readmission, Subsequent Ligament Surgery and Subsequent TKA.
| CMS Complications |
90-day Readmission |
Subsequent Ligament Surgery |
Subsequent TKA |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Odds Ratio | 95% Confidence Interval | P-value | Odds Ratio | 95% Confidence Interval | P-value | Odds Ratio | 95% Confidence Interval | P-value | Odds Ratio | 95% Confidence Interval | P-value | |
| CCIa | 1.4 | (1.0, 1.8) | 0.04 | |||||||||
| Obesity | 4.2 | (1.8, 9.5) | <.001 | |||||||||
| Ambulatory Surgical Centerb | 0.05 | (.004, .6) | 0.007 | |||||||||
| Hospital-based Outpatient Departmentb | 0.04 | (.004, .4) | 0.01 | |||||||||
| Knee Dislocation Diagnosis | 4 | (1.7, 9.5) | <.001 | 2.5 | (1.2, 5.2) | 0.01 | ||||||
| Concomitant Meniscectomy | 9.1 | (1.4, 67.0) | 0.02 | |||||||||
| Prior Placement of External Fixator | 1.3 | (1.1–1.7) | 0.02 | |||||||||
| Concomitant Neuroplasty | 1.5 | (1.37–1.9) | <.001 | |||||||||
Odds ratio for every point increase in CCI score.
Reference group is inpatient surgical setting.
4. Discussion
This study attempted to define the epidemiology and identify trends of multi-ligament knee reconstruction through the analysis of a national insurance database. 588 patients were identified undergoing multi-ligament knee reconstruction from 2007 to 2016. The purpose of this study was to identify factors associated with 90-day complications and readmissions after multi-ligament knee reconstruction as well as with subsequent ligament surgery and TKA.
The mean age of patients sustaining multi-ligament knee injury reported in the literature ranges from the second to fourth decade of life and the majority of patients are male.2,7,13,14 The association between these injuries and high-energy trauma and sports, as well as greater male participation in these activities likely explain these findings. That this study agrees with these trends previously identified through single center experiences suggests that they may be nationally representative. What is absent in the literature is the percentage of patients undergoing multi-ligament knee surgery with a history of traumatic knee dislocation. Not all multi-ligamentous knee injuries are knee dislocations and further research stratified by mechanism of injury may be helpful as this study demonstrated that knee dislocation diagnosis was strongly associated with readmission and subsequent surgery.15
The readmission rate of 8.3% was higher than that of previous literature analyzing a statewide database.1 The difference may reflect attritional losses to follow-up of patients who may have had surgery in one state but were readmitted in another and may also reflect the national heterogeneity in complexity of and complications following surgery that may require readmission. Nevertheless, a diagnosis of dislocation and surgery taking place in an inpatient setting are strong predictors of readmission, which is echoed by previous studies.1 In fact, we showed that patients undergoing surgery in a hospital-based outpatient department and ambulatory surgical center are 20 times less likely to be readmitted relative to those undergoing surgery in inpatient settings. This may be a function of patient selection as certain patients with high risk comorbidities, such as renal disease, electrolyte imbalances and blood dyscrasis, are more likely to require an inpatient setting of care peri-operatively and are also more likely to be readmitted.1 In this study, the mean Charlson Comorbidity Index score of the inpatient group was significantly greater than that of the outpatient group (1.38 vs .43, p < .01). Of note, comorbidity burden was adjusted for in our regression analysis of readmission and was not found to be a significant predictor.
This study sheds new light on the risk of subsequent ipsilateral ligament surgery following multi-ligament knee surgery whereas previous studies relying on coding are limited in their determination of laterality of index and subsequent procedures.1,7 Reported rates of revision surgery range from 9% to 36%.15,16 In a study of 133 patients undergoing multi-ligament knee surgery, Cook et al. did not identify repair versus reconstruction or graft choice (allograft or autograft) to be factors associated with the need of revision surgery.15 They did, however, find a significant association between patients with a Schenck classification of KD IV and risk for revision surgery, suggesting that injury severity may play an important role. This is further corroborated by this study which demonstrated an association between likely surrogates of injury severity and complexity (ie. diagnosis of knee dislocation, previous placement of external fixator, concomitant neuroplasty) and increased likelihood of subsequent ligament surgery. Future research targeted towards indications and risk factors for revision in the setting of high risk multi-ligament knee injuries may help improve our understanding.
While there are certain advantages of querying a national insurance cohort that is likely representative of the US population in terms of patient, surgeon and hospital-related factors, there are numerous limitations worth mentioning. We were unable to obtain information on the timing and mechanism of injury, concomitant injuries both related and unrelated to multi-ligament knee injury, initiation of physical therapy, surgical timing and whether intervention was staged or not. On analysis of subsequent ipsilateral ligament surgery, we excluded any subsequent surgeries performed within 6 months of index surgery to account for deliberate staging of ligament reconstruction.8,9
This study may underreport the true incidence of multi-ligament reconstructions. To standardize the follow-up period of our sample, we included only patients who were active within the insurance database at a minimum of 6 months before and 18 months after the index surgery. There may be a subset of relevant patients who switched insurances or dropped out of the insurance dataset within this time period. This study defined a multi-ligament procedure by a combination of CPT and ICD-9 codes designating both a cruciate and collateral ligament procedure which does not coincide with clinically relevant classifications either by the direction of preceding knee dislocation or by the number of ligaments requiring operative repair or reconstruction. Moreover, the use of an administrative dataset precludes more granular analysis of the heterogenous group of multi-ligament procedures and is affected by coding error and variation.17 Lastly, this study could not account for surgeon experience and hospital volume which may play a role in outcomes.
5. Conclusions
This study provides new information from a national database on contemporary trends on multi-ligament reconstruction and factors associated with 90-day complications, readmissions, and subsequent ligament surgery and total knee arthroplasty. Knowledge of these factors may help future surgeons modify their practice, surgical techniques and patient education to decrease the risk of these complications. Understanding common factors predisposing this population to adverse events following surgery may assist in improved outcomes. Future research should be directed towards patients at higher risk for these complications.
Declaration of competing interest
We, the authors, have no conflicts of interest to disclose with regards to this study.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.jor.2020.01.030
Appendix A. Supplementary data
The following is the supplementary data to this article:
References
- 1.Wilson S.M., Mehta N., Do H.T., Ghomrawi H., Lyman S., Marx R.G. Epidemiology of multiligament knee reconstruction. Clin Orthop Relat Res. 2014;472:2603–2608. doi: 10.1007/s11999-014-3653-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Arom G.A., Yeranosian M.G., Petrigliano F.A., Terrell R.D., McAllister D.R. The changing demographics of knee dislocation: a retrospective database review. Clin Orthop Relat Res. 2014;472:2609–2614. doi: 10.1007/s11999-013-3373-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Alentorn-Geli E., Lazarides A.L., Utturkar G.M. Factors predictive of poorer outcomes in the surgical repair of multiligament knee injuries. Knee Surg Sports Traumatol Arthrosc. 2019;27:445–459. doi: 10.1007/s00167-018-5053-9. [DOI] [PubMed] [Google Scholar]
- 4.Bagherifard A., Jabalameli M., Ghaffari S. Short to mid-term outcomes of single-stage reconstruction of multiligament knee injury. The archives of bone and joint surgery. 2019;7:346–353. [PMC free article] [PubMed] [Google Scholar]
- 5.King A.H., Krych A.J., Prince M.R., Sousa P.L., Stuart M.J., Levy B.A. Are meniscal tears and articular cartilage injury predictive of inferior patient outcome after surgical reconstruction for the dislocated knee? Knee Surg Sports Traumatol Arthrosc. 2015;23:3008–3011. doi: 10.1007/s00167-015-3671-z. [DOI] [PubMed] [Google Scholar]
- 6.Levy N.M., Krych A.J., Hevesi M. Does age predict outcome after multiligament knee reconstruction for the dislocated knee? 2- to 22-year follow-up. Knee Surg Sports Traumatol Arthrosc. 2015;23:3003–3007. doi: 10.1007/s00167-015-3750-1. [DOI] [PubMed] [Google Scholar]
- 7.Brautigan B., Johnson D.L. The epidemiology of knee dislocations. Clin Sports Med. 2000;19:387–397. doi: 10.1016/s0278-5919(05)70213-1. [DOI] [PubMed] [Google Scholar]
- 8.Jiang W., Yao J., He Y., Sun W., Huang Y., Kong D. The timing of surgical treatment of knee dislocations: a systematic review. Knee Surg Sports Traumatol Arthrosc. 2015;23:3108–3113. doi: 10.1007/s00167-014-3435-1. [DOI] [PubMed] [Google Scholar]
- 9.Levy B.A., Dajani K.A., Whelan D.B. Decision making in the multiligament-injured knee: an evidence-based systematic review. Arthroscopy. 2009;25:430–438. doi: 10.1016/j.arthro.2009.01.008. [DOI] [PubMed] [Google Scholar]
- 10.Levy B.A., Dajani K.A., Morgan J.A., Shah J.P., Dahm D.L., Stuart M.J. Repair versus reconstruction of the fibular collateral ligament and posterolateral corner in the multiligament-injured knee. Am J Sports Med. 2010;38:804–809. doi: 10.1177/0363546509352459. [DOI] [PubMed] [Google Scholar]
- 11.Lian J., Patel N.K., Nickoli M. Obesity is associated with significant morbidity after multiligament knee surgery. J Knee Surg. 2019 Mar 1 doi: 10.1055/s-0039-1681027. [DOI] [PubMed] [Google Scholar]
- 12.Ridley T.J., Cook S., Bollier M. Effect of body mass index on patients with multiligamentous knee injuries. Arthroscopy. 2014;30:1447–1452. doi: 10.1016/j.arthro.2014.05.035. [DOI] [PubMed] [Google Scholar]
- 13.Rios A., Villa A., Fahandezh H., de Jose C., Vaquero J. Results after treatment of traumatic knee dislocations: a report of 26 cases. J Trauma. 2003;55:489–494. doi: 10.1097/01.TA.0000043921.09208.76. [DOI] [PubMed] [Google Scholar]
- 14.de Loes M., Dahlstedt L.J., Thomee R. A 7-year study on risks and costs of knee injuries in male and female youth participants in 12 sports. Scand J Med Sci Sports. 2000;10:90–97. doi: 10.1034/j.1600-0838.2000.010002090.x. [DOI] [PubMed] [Google Scholar]
- 15.Cook S., Ridley T.J., McCarthy M.A. Surgical treatment of multiligament knee injuries. Knee Surg Sports Traumatol Arthrosc. 2015;23:2983–2991. doi: 10.1007/s00167-014-3451-1. [DOI] [PubMed] [Google Scholar]
- 16.Stannard J.P., Brown S.L., Farris R.C., McGwin G., Jr., Volgas D.A. The posterolateral corner of the knee: repair versus reconstruction. Am J Sports Med. 2005;33:881–888. doi: 10.1177/0363546504271208. [DOI] [PubMed] [Google Scholar]
- 17.Patel A.A., Singh K., Nunley R.M., Minhas S.V. Administrative databases in orthopaedic research: pearls and pitfalls of big data. J Am Acad Orthop Surg. 2016;24:172–179. doi: 10.5435/JAAOS-D-13-00009. [DOI] [PubMed] [Google Scholar]
- 18.Hohmann E., Glatt V., Tetsworth K. Early or delayed reconstruction in multi-ligament knee injuries: a systematic review and meta-analysis. Knee. 2017 Oct 1;24(5):909–916. doi: 10.1016/j.knee.2017.06.011. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.