703 posters,  63 sessions,  7 topics,  1978 authors 
ePostersLive® by SciGen® Technologies S.A. All rights reserved.

4930
Development in bias and sample size research on postoperative pain management after hip and knee arthroplasty
Session: MP-01a
Thurs, April 19, 8-9:45 am
Shubert (Shubert Complex), 6th floor

Please note, medically challenging cases are removed three months after the meeting and scientific abstracts after three years.

Poster Presenter
Affiliations

Rate

No votes yet

Development in bias and sample size for research on postoperative pain management after hip and knee arthroplasty

Introduction: 

Bias (systematic error) and small trial sample size (random error) may induce imprecise and exaggerated treatment effects in randomized clinical trials (RCTs) [1,2]. Increasing adherence to CONSORT-guidelines and Cochrane Collabora-tion bias recommendations in the past decades may have improved these issues.

Materials and Methods:

Data were retrieved from two systematic reviews of RCTs on pain management after total hip and knee arthroplasty [3,4]. Co-primary outcomes were risk of bias and trial sample size developments over time. We calculated bias scores from 0-14 based on Cochrane’s seven bias domains (0=low; 1=unclear, 2=high). Development was evaluated in run and control charts. These are point-and-line graphs of development over time. A hori-zontal line indicates the median or mean. Charts test for non-random variation and process instability in data. We used three tests for this: 1) The curve having unusually long runs of data points on  one side of the median. 2) The curve crossing the median unusually few times. 3) Data points outside control limits for natural process variation (grey area) [5].We also compared adequate reporting in trials published between 1990-1999 and 2010-2016.

Results:

We included 171 trials published between 1989-2016. The summarized risk of bias decreased, mainly due to better randomization and allocation concealment (Figure 1). Visual inspection suggested an on-going improvement that started around 2007. Trial sample size did not change significantly. Adequate reporting increased when comparing trials published in 1990-1999 and 2010-2016 for all bias domains (Figure 2).

Discussion and Conclusions:

In decision-making regarding analgesic regimens, the evidence from especially older trials, should be interpreted with care. Optimization of study quality should be prioritized in planning of new trials. 1. Risk of bias decreased from 2007. 2. The randomization and allocation process has improved. 3. Sample size has not increased.

 

Enter Poster ID (e.gGoNextPreviousCurrent