Impact of age and comorbidity burden on mortality and major complications in older adults undergoing orthopaedic surgery: an analysis using the Japanese diagnosis procedure combination database

BACKGROUND The purpose of this study was to examine how complications in older adults undergoing orthopaedic surgery vary as a function of age, comorbidity, and type of surgical procedure. METHODS We abstracted data from the Japanese Diagnosis Procedure Combination database for all patients aged ≥ 50 who had undergone cervical laminoplasty, lumbar decompression, lumbar arthrodesis, or primary total knee arthroplasty (TKA) between July 1 and December 31 in the years 2007 to 2010. Outcome measures included all-cause in-hospital mortality and incidence of major complications. We analyzed the effects of age, sex, comorbidities, and type of surgical procedure on outcomes. Charlson comorbidity index was used to identify and summarize patients' comorbid burden. RESULTS A total of 107,104 patients were identified who underwent cervical laminoplasty (16,020 patients), lumbar decompression (31,605), lumbar arthrodesis (18,419), or TKA (41,060). Of these, 17,339 (16.2%) were aged 80 years or older. Overall, in-hospital death occurred in 121 patients (0.11%) and 4,448 patients (4.2%) had at least one major complication. In-hospital mortality and complication rates increased with increasing age and comorbidity. A multivariate analysis showed mortality and major complications following surgery were associated with advanced age (aged ≥ 80 years; odds ratios 5.88 and 1.51), male gender, and a higher comorbidity burden (Charlson comorbidity index ≥ 3; odds ratio, 16.5 and 5.06). After adjustment for confounding factors, patients undergoing lumbar arthrodesis or cervical laminoplasty were at twice the risk of in-hospital mortality compared with patients undergoing TKA. CONCLUSIONS Our data demonstrated that an increased comorbid burden as measured by Charlson comorbidity index has a greater impact on postoperative mortality and major complications than age in older adults undergoing orthopaedic surgery. After adjustment, mortality following lumbar arthrodesis or cervical laminoplasty was twice as high as that in TKA. Our findings suggest that an assessment of perioperative risks in elderly patients undergoing orthopaedic surgery should be stratified according to comorbidity burden and type of procedures, as well as by patient's age.