Prevalence of anemia and association with outcome in joint arthroplasty - is there a difference between primary and revision cases?

purposeAnemia has been shown to be a modifiable pre-operative, patient factor associated with outcome following arthroplasty. The aims of this retrospective study were to (1) ascertain the prevalence of preoperative anemia in patients undergoing primary and revision hip and knee arthroplasty at a tertiary referral center and (2) to test the association with outcome and whether it differs between primary and revision cases.

methodsAll hip and knee primary and revision arthroplasties performed at a Canadian academic, tertiary-care, arthroplasty center between 2012 and 2017 were included in this study. The study group consisted of 5944 patients, of which 5251 were primary Total Hip and Knee Arthroplasties or Hip Resurfacings and 693 were revision arthroplasties (65% hip revisions/35% knee revisions). Anemia was classified as per WHO definition (hemoglobin < 130 g/L for men and < 120 g/L for women). All anemic patients were grouped into mild, moderate or severe anemia. Length-of-stay, perioperative transfusion-rate, 90-day readmission, overall complication rate and reoperation rates were recorded. The effect of preoperative anemia and the effect of severity of the anemia was evaluated through multivariable regression analysis controlling for relevant covariates.

resultsPreoperatively, 15% (786/5251) of the primary patients and 47% (322/693) of the revision arthroplasty patients were anemic preoperatively. Anemic revision patients were 3.1 times more likely (95% CI: 1.47-6.33) to obtain blood transfusions during the hospital stay, compared to a 4.9 times higher risk in primary patients. The odds ratio to sustain any postoperative complication if anemic was 1.5 times higher (95% CI: 0.73-3.16) in revision patients and 1.7 in primary cases. In addition, the 90-day readmission rate among both groups was 1.6 times higher in anemic patients. Furthermore, anemic revision patients had a 5.3 days longer length of stay (95% CI: 2.63-7.91), compared to only 1 additional day in anemic primary patients (95% CI: 0.69-1.34).

conclusionIn this study cohort, the prevalence of anemia in patients awaiting revision arthroplasty was 3 times higher (46.6%) than in primary arthroplasty patients (18.7%). Preoperative anemia was associated with similarly, inferior outcomes in both groups. To reduce postoperative complications and the "burden" associated with anemia, these findings strongly recommend optimizing the preoperative hemoglobin in all arthroplasty patients. However, revision patients are affected more frequently, and particular attention must therefore be taken to this growing group in the future. LEVEL OF EVIDENCE: Level III.