Table 1.

Descriptive summary of 15 studies on the proportion of undernutrition among HIV positive Adults in Ethiopia included for meta-analysis, 2006–2016

Study setting	study year	Study Design	Age	Response rate (%)	Event	Sample Size (N)	Some potential biases in their methods
Humera Hospital	2012	Cross-sectional	≥20	100	159	376	Source population and study population were not defined, outcome variable was not defined, and model fitness was not checked.
Dembia District Health Centers	2015	Cross-sectional	≥18	100	105	452	Multicollinearity was not assessed and pretest was not done.
University of Gondar Referral Hospital	2007	Cross-sectional	≥15	99.7	92	331	Multicollinearity and model fitness were not assessed.
University of Gondar Referral Hospital	2016	Cross-sectional	≥18	99.4	58	319	Training for data collectors was not given and Multicollinearity was not assessed.
Mettema Hospital	2013	Cross-sectional	≥18	99.5	131	378	Multicollinearity was not assessed.
Bahir Dar Referral Hospital	2009	Cross-sectional	≥18	100	104	408	Standardized tool was not used, outcome variable was not defined, multicollinearity was not assessed and model fitness was not checked.
Addis Ababa	2006	Cross-sectional	≥18	100	128	153	Relatively has small sample size, confounders were not controlled in the statistical analysis, standardized tool was not used, pretest was not done, and training for data collection was not given.
Jimma University specialized Hospital	2014	Cross-sectional	≥18	95.6	51	109	Relatively has small sample size, non-probable sampling technique was employed to select study participants, training for data collectors was not given, and model fitness and multicollinearity were not assessed.
Jimma University specialized Hospital	2009	Cross-sectional	≥18	100	63	319	Training for data collectors was not given, model fitness was not assessed, outcome variable was not defined, and multicollinearity was not assessed.
Nekemte Referral Hospital	2013	Cross-sectional	≥15	100	115	423	The study population and source population were not clearly defined, outcome variable was not defined, and training for data collectors was not given.
Dilla University Referral Hospital	2012	Cross-sectional	≥18	100	64	520	Inclusion and exclusion criteria were not mentioned, pretest was not done, training for data collectors was not given, and model fitness and multicollinearity were not assessed.
Butajira Hospital	2014	Cross-sectional	≥18	100	77	305	Outcome variable was not defined.
Health Facilities of Hosana Town	2014	Cross-sectional	≥18	97.1	103	330	Model fitness and multicollinearity were not assessed.
In Hawassa Health Institutions	2014	Cross-sectional	20–50	99.4	119	719	Population who are not willing to participate were excluded from the study, pregnant women were included in the study and assessed using BMI this will result bias results, and model fitness and multicollinearity were not assessed.
Wolaita Sodo Teaching and Referral Hospital	2016	Cross-sectional	≥18	96.3	133	500	Exclusion criteria was not mentioned, and model fitness and multicollinearity were not assessed.