Invited Paper

A Study of the Approaches to Treating Missing Data in the Data Collected with the WHODAS 2.0 Scale

10.4274/atfm.galenos.2019.30075

  • Duygu Sıddıkoğlu
  • Beyza Doğanay Erdoğan
  • Derya Gökmen
  • Şehim Kutlay

Received Date: 09.01.2019 Accepted Date: 14.01.2019 J Ankara Univ Fac Med 2019;72(1):1-12

The aim of this study was to evaluate whether missing data rates are indicated in articles which applied World Health Organization Disability Assessment Schedule 2.0 scale (WHODAS 2.0); and to determine if missing data analysis were done and how it was conducted.

We conducted a systematic literature review on articles applying WHODAS 2.0 about strategies used for missing data analysis from November 4, 1999, to April 6, 2018. We evaluated whether the rates of missingness is given by domain and item, how the handling missing data strategies proposed in manual for WHODAS 2.0 taken into consideration and other approaches for missing data analysis.

A total of 91 abstracts were reviewed and 82 articles chosen for full text review. Thirty-two articles were included in the final evaluation. Below 30% of articles gave the rate of overall missingness and rates by domain. Only 9.38% of articles gave missing rate for some items that have the highest missing rate. In 50% of articles, the approaches proposed in the WHODAS 2.0 manual (single imputation and multiple imputation) were performed in different ways.

When analyzing the approaches for missing data analysis among reviewed articles, it is difficult to get a consistent method. Authors implemented missing data approaches proposed in the WHODAS 2.0 application manual in different ways. There is no common sense for the critical rates of missingness. In addition to not applicable items related to work/school life activities (D5.8-D5.11) which is mentioned in the manual, a systematic missingness was observed in D4.5 “sexual activities” item.

This study is the first study on missing data in WHODAS 2.0 literature and it can be used as guideline for reflecting possible missing data to researchers who will apply WHODAS 2.0 and also will be useful as a preliminary study on methodological studies on missing data.

Keywords: WHODAS 2.0, Missing Data, Systematic Review

Full Text (Turkish)