gms | German Medical Science

Objectives: The primary aim of our study was to compare two different documentation ways: external documentation (quality assurance) vs. internal documentation after primary hip arthroplasty. Secondary aims were to evaluate the changes in the complication rates across the period under study and the relevance of patient's gender, duration of operation, duration of hospital stay and patient age with the complication rate.

Methods: Data concerning primary total hip arthroplasties conducted in our clinic from 2005 until 2014 were prospectively collected in two different ways: internal documentation and external documentation (for quality assurance). Statistical analyses using Pearson's chi-squared test or Fisher's exact test were performed to compare internal to external data, and complications observed during the years 2005-2009 and 2010-2014. The Mann-Whitney U test was used to compare the continuous outcomes and age of patients in the aforementioned periods. The complications of the year 2004 were excluded in order to create two 5-year periods. Furthermore, binary logistic regression analyses were performed (dependent variable: occurrence of complication, independent variables: gender, normalized duration of operation, normalized length of hospital stay and normalized patient age). Statistical significance was set at a p-value of 0.05.

Results and Conclusion: The total number of primary hip arthroplasties was 3,395. The median age of the patients, who sustained a complication, was 70 years (range: 20-90) and 32.7% of them were male. The number of patients experiencing a postoperative complication was 262, after exclusion of patients experiencing postoperatively ambiguously defined changes in the mobility. The median postoperative length of stay in hospital of patients experiencing a complication was 15 days (range: 1-72) and the median duration of operation was 96 minutes (range:39-298).

A statistical significant difference between the two datasets (internal vs. external) was detected in the "inability to walk at discharge" (p<0.0001) and the occurrence of "postoperative wound infection" (p=0.0038).

Furthermore, significant reduction of the complication rate was observed during the period 2010-2014 according to the internal and external documented "total complication rate" (p<0.0001). A significant difference was found in the following complication rates of both documentation methods (internal and external): "hematoma and postoperative hemorrhage", "cardiovascular complications" and "reintervention" (p<0.01).

Finally, binary logistic regression analysis was performed, removing insignificant independent variables. The variable "normalized length of hospital stay" more commonly predicted independently the occurrence of complications.

Further improvements should be made in the documentation processes. Finally, assumptions could be made about the role of surgical experience gained through years and the impact of implementation of a quality assurance program on observed reduction of complications.