Intimate partner violence (IPV): The validity of an IPV screening instrument utilized among pregnant women in Tanzania and Vietnam

Research output: Contribution to journalJournal articleResearchpeer-review


Intimate partner violence (IPV) is a global problem that affects one-third of all women. The present study aims to develop and determine the validity of a screening instrument for the detection of IPV in pregnant women in Tanzania and Vietnam and to determine the minimum number of questions needed to identify IPV.
An IPV screening instrument based on eight questions was tested on 1,116 Tanzanian and 1,309 Vietnamese women who attended antenatal care before 24 gestational weeks. The women were re-interviewed during their 30th-34th gestational week where the World Health Organization (WHO) IPV questionnaire was used as the gold standard. In all, 255 combinations of eight different questions were first tested on the Tanzanian study population where sensitivity, specificity, positive predictive value, negative predictive value and accuracy were calculated. In the evaluation of the performance of the question combinations, different IPV types and the frequency of abusive acts were considered. The question combinations that performed best in Tanzania were subsequently evaluated in the Vietnamese study population.
In Tanzania, a combination of three selected questions including one question on emotional IPV, one on physical IPV and one on sexual IPV was found to be most effective in identifying women who are exposed to at least one type of IPV during pregnancy (sensitivity = .80; specificity = .74). The performance of the identified combination was slightly less effective in Vietnam (sensitivity = .74; specificity = .68). Focusing on different IPV types, the best performance was found for exposure to physical IPV in both Tanzania (sensitivity = .93; specificity = .70) and Vietnam (sensitivity = .96; specificity = .55). In both countries, the sensitivity increased with the frequency of abuse whereas the specificity decreased.
By asking pregnant women three simple questions we were able to identify women who were exposed to IPV during pregnancy in two different countries. The question combination performed best in assessing physical IPV where it identified 93% and 96% of Vietnamese and Tanzanian women, respectively, who were exposed to physical IPV.
Original languageEnglish
Article numbere0190856
JournalP L o S One
Issue number2
Publication statusPublished - 1 Feb 2018

Number of downloads are based on statistics from Google Scholar and

No data available

ID: 193532872