Incidence, Outcomes, and Prediction of Postoperative Urinary Retention After a Nonurologic Procedure

Abstract

Purpose: To develop a prognostic model to estimate postoperative urinary retention (POUR) after lower limb arthroplasty.

Methods: One thousand two hundred twenty patients underwent 1,374 joint replacement operations (812 knees and 562 hips) between December 2008 and May 2014. Detailed variables were collected. A multivariable logistic regression model was used to identify the independent predictors for POUR. Boot strapping and stepwise elimination was used to design a predictive nomogram.

Results: There were 124 incidents of POUR (9.02%) in 118 patients (90 knee, 34 hip, P = 0.001). On univariate analysis, patients who developed POUR were older (P < 0.001), had higher American Association of Anesthesiology scores (P = 0.007), underwent knee replacement (0.001), were obese (body mass index > 35) (P = 0.04), and were hypertensive (P = 0.029), with a history of benign prostatic hyperplasis (BPH) (P < 0.001) or neurologic disorders (P = 0.024). On multivariable analysis, age (60 to 69 years, P = 0.023, 70 to 79 yrs P = 0.008, >80 years P = 0.003), knee replacement (P = 0.014), and history of BPH (P = 0.013) were the independent predictors of POUR. A score was assigned to each predictor (total = 31). The C-index was 0.65. There were three risk categories as follows: 0 to 50, 51 to 85, and 86+ points resulting in 3.3%, 7.2%, and 14.0% risk of retention, respectively.

Discussion: This nomogram reliably predicts the risk of POUR in patients undergoing hip and knee arthroplasties and may help planning preoperative interventions to decrease the risk of this complication.

Conflict of interest statement

None of the following authors or any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Dr. Abdul-Muhsin, Dr. Jakob, Dr. Cha, Dr. Zhang, Dr. Schwartz, Dr. Navaratnam, Dr. Khan, and Dr. Humphreys.

Figures

Figure 1

Figure demonstrating the predictive nomogram to predict postoperative urinary retention after weight-bearing joint replacement surgery.

Figure 2

Figure demonstrating the calibration plot to assess the accuracy of urinary retention nomogram after joint replacement surgery. The “Ideal” line represents perfect prediction because the predicted probabilities equal the observed probabilities. The “Apparent” line represents the calibration using our sample. The “Bias Corrected” line is derived via resampling. The closer the bias corrected line to the ideal line the more accurate is the prediction within that specific range. When the line is below the ideal line, it is overprediciting the outcome and when it is above the ideal line, it is underpredicting the outcome.