Do Cultures Obtained During Primary THA Predict the Likelihood of Revision?

Background: There can be unexpectedly positive culture results during elective hip arthroplasty, but the degree to which these are associated with an increased risk of subsequent premature revision is not known.

Question/purpose: Are unexpectedly positive culture results obtained during elective THA associated with an increased likelihood of revision within 5 years of the procedure?

Methods: Between March 2007 and March 2011, the hip unit at our institution performed elective primary THA in 829 patients. We systematically collected three samples in 52% (428 of 829) of the interventions. Of those, 26 patients were excluded because of sampling errors; 94% (402 of 428) had samples that were collected systematically and were eligible for the study. We only considered one hip randomly in bilateral procedures (4% [15 of 428]); patients presenting with acute (< 3 months) periprosthetic joint infection undergoing open debridement (4% [16 of 402]) and patients who died before 5 years of follow-up (2% [seven of 402]) were excluded from the study, leaving 91% (364 of 402) eligible for analysis in this retrospective study of a previous prospective trial. No patient included in the final analysis was lost to follow-up within 5 years from the index surgery. The patient group consisted of 52% (188 of 364) women, with a mean ± SD age of 64.8 ± 13.9 years.

Results: Positives culture results were associated with a higher risk of revision within 5 years of the index surgery. The proportion of revision surgery was higher in the group with positive culture results than in those with negative results (10% [eight of 77] versus 2% [seven of 290]; p = 0.01). The difference was mainly attributable to a higher proportion of aseptic loosening in those with positive culture results than in those with negative results (8% [six of 74] versus 1% [four of 290]; p = 0.01). After a multivariable analysis, the only independent variable associated with 5-year revision surgery was the presence of positive results during THA (odds ratio 4.9 [95% confidence interval 1.72 to 13.99]).

Conclusion: Our findings suggest that bacterial contamination during THA is associated with an increased likelihood of early revision. This higher risk of revision is mainly because of presumed aseptic loosening; thus, efforts should focus on the need to rule out infection. These results not only open new questions that should be answered in new prospective and well-designed studies, but also may help to better select patients to obtain a more favorable outcome after THA.

Level Of Evidence: Level III, therapeutic study.