Summary

For the majority of patients undergoing shoulder arthroplasty, routine postoperative laboratory analyses do not change acute postoperative management.

Abstract

Background

Shoulder arthroplasty is routinely performed in the inpatient setting. While shoulder arthroplasty is associated with low patient morbidity, in many institutions, postoperative laboratory studies are routinely obtained, potentially increasing healthcare costs without providing significant benefit. The purpose of this study was to determine the utility of obtaining these tests in the postoperative setting. The authors hypothesized that routine laboratory analysis on the first post-operative day (POD) following shoulder arthroplasty would have minimal to no impact on changing the acute inpatient management of the patient.

Methods

1063 consecutive patients undergoing inpatient total shoulder arthroplasty (TSA), reverse shoulder arthroplasty (RSA), or hemiarthroplasty between 2010 and 2014 by 2 senior surgeons were analyzed. Patient demographics, comorbidities, and length of stay (LOS) were assessed. Preoperative laboratory values were compared with POD-1 values. Intraoperative estimated blood loss (EBL) and transfusion rates were analyzed. Statistical analysis was performed with Mann-Whitney U tests, and chi-square or fischer’s exact tests where appropriate, with P<0.05 considered significant.

Results

After meeting inclusion criteria, 954 patients (460 males, 494 females; average age 65.6±11.1 years) undergoing 415 TSA, 434 RSA, and 105 hemiarthroplasties were included. The average operative time, EBL, and LOS were 114.23±25.05 min, 255.35±169.16 mL, and 51.47±24.45 hours, respectively. The average preoperative to postoperative decrease in HGB was 2.88±1.17 g/dL (P<0.0001). Thirty-nine patients (4.1%) required postoperative blood transfusions at an average of 1.59±1.00 days after surgery. Subgroup analysis revealed significantly lower average preoperative HGB (P<0.001), higher EBL (P<0.0001), and longer LOS (P<0.0001) in the 39 patients requiring transfusion compared to the 915 patients not transfused. Transfusion rates were highest following RSA (7.4%), followed by hemiarthroplasty (1.9%) and TSA (1.2%, P<0.0001). Independent risk factors for transfusion were age >72 years, female sex, and the use of a drain. ROC analysis demonstrated preoperative HGB of 13.75 g/dL, and hematocrit value of 42.85%, as 90% sensitivity cutoffs for transfusion.

Conclusions

For the majority of patients undergoing shoulder arthroplasty, routine postoperative laboratory analyses do not change acute postoperative management. Certainly, the decision to obtain postoperative labs should be considered on a patient-specific basis.