The present study is part of a larger study of methodological research conducted in an exploratory sequential mixed method in two qualitative stages (designing obstetrics triage satisfaction tool) and quantitative (examining psychometric properties of obstetrics triage satisfaction tool).
Inclusion criteria included all recipients of obstetrics triage services at Shahid Beheshti University of Medical Sciences teaching hospitals in Tehran. Sampling was done from June to November 2019.
Waltz's four-step tool design was done. The deductive inductive method was used to extract the items of obstetrics triage satisfaction assessment. To illustrate the concept of obstetrics triage satisfaction and its dimensions, the qualitative method of content analysis was used, and then a literature review was performed. Semi-structured interviews were conducted with 23 participants including individuals referring to obstetrics triage, triage personal, and key obstetrics informants. Participants were selected with maximum diversity in terms of age, education, and field of study. The interview started with the general question ‘How was the obstetrics triage service? and continued with more specific questions. The duration of each interview was 15–20 minutes. The number of interviews continued until information saturation. Data analysis was done, as data was collected, based on steps proposed by Graneheim and Lundman(19, 20). After each interview, the interviews were transcribed after listening several times; they were typed and then entered into the MAXQDA10 software. The transcripts of the interviews were coded several times of re-reading by the researcher using an open coding system to produce the first categories. For this purpose, the texts of the interviews were first divided into semantic units, and then different codes were compared based on their similarities and differences and categorized into categories. Then, a comprehensive review of the studies was done using the deductive approach and key terms of triage, obstetric or midwifery triage, satisfaction, questionnaire, scale, index, and checklist were searched in PubMed, Proquest, Science Direct, Ovid, Google Scholar, Magiran, Iranmedex, Irandoc databases. Inclusion criteria included related Persian or English articles, regardless of the time limit. Articles were then reviewed and items related to obstetric triage quality not included in the phrase pool were added to the main items of the questionnaire. The initial tool was then prepared for validation.
At this stage, face validity, content validity, and construct validity were performed. Reliability and validity were also evaluated.
Face validity: Face validity was assessed by qualitative and quantitative methods and content validity by qualitative and quantitative methods respectively.
In the qualitative face validity, the items of the questionnaire were evaluated in terms of difficulty level, appropriateness, and ambiguity of items by five referring to obstetric triage. In quantitative formal validity, the impact index was used. Fifteen participants were asked to rate the importance of each item on the five-point Likert scale. If the impact score is equal to or greater than 1.5, the item is recognized as suitable for subsequent analysis(21).
Content validity: To determine the content validity, qualitative and quantitative methods were used. In reviewing the qualitative content validity, the researcher asked four instrumentalists, emergency gynecologists, obstetricians, and midwives to comment on clarity, simplicity, grammar use, appropriate words, placement of items in appropriate self-assessment, and appropriate scoring in written form(21, 22). To calculate the content validity ratio (CVR) quantitatively, 15 experts (Emergency Medicine, Gynecology, Reproductive Health, Midwifery) were asked to rank each item based on the three-point scale, "necessary, useful but not necessary, and not necessary” to check the necessity of the item. The expressions whose numerical CVR values above 0.49 were retained based on the Lawshe table (for 15 evaluates) to determine the minimum CVR(23).
In investigating the content validity index, the relevance of the items to the purpose of the research was evaluated from specialists (emergency medicine, gynecology, reproductive health, midwifery). To determine this index, 15 experts were asked to examine the relevance and adequacy of the items based on the four-point Likert scale. In this study, two indices, Item Content Validity Index (I-CVI) and Scale Content Validity Index (S-CVI) were calculated. The I-CVI represents the ratio of agreement on the relevance of each item. It is obtained by dividing the number of experts who rated the relevance or appropriateness of item 3 or 4 by the total number of experts. The validity index score above 0.79 is appropriate, 0.79 − 0.7 is questionable and needs to be revised and a score below 0.7 is unacceptable and should be deleted (22, 24). The S-CVI acceptance criterion among toolmakers is 0.9 as the highest criterion and 0.8 as the lower content acceptance threshold (22).
Construct validity Exploratory factor analysis was used to determine the construct validity. In this study, according to the number of 10 items, 200 questionnaires were completed by the individuals referring to obstetrics triage considering the loss probability (25). The convenience method was used for sampling. The questionnaires were completed by those referring to obstetrics triage of hospitals in North, South, East, and Western areas affiliated to Shahid Beheshti University of Medical Sciences. To determine the adequacy of the sample, the Kaiser-Meyer-Olkin (KMO) test was used. The latent factors were extracted using maximum likelihood analysis and Promax rotation assuming the independence of the factors. In this model, two factors with eigenvalues higher than one were extracted. The cutoff point of 0.4 was considered, the minimum factor loading required for each statement to retain it.
The reliability of the tool was assessed by internal consistency and Cronbach's alpha coefficient. An alpha coefficient above 0.7 was considered acceptable (26). To check the consistency of the obstetrics triage satisfaction questionnaire, the satisfaction questionnaire was completed by a sample of individuals (15 recipients of obstetrics triage service) in two stages with two weeks’ interval and then scores obtained in two stages were calculated using intra-cluster correlation coefficient (ICC) test. The intra-cluster correlation coefficient of 0.7 and above denotes the optimal consistency.
A ceiling effect is said to occur when a high proportion of subjects in a study have maximum scores on the observed variable. This discriminates among subjects among the top end of the scale impossible, and the floor effect arises when a high proportion of subjects select the responses on the lower limit. The index should be less than 20% to include all criteria and show changes over time. In this study, the ceiling effect and floor effect was calculated for the total score of the questionnaire and the score of all subscales as a percentage to evaluate the scale discrimination and response distribution(27).
Ease and feasibility
Ease of use, ease of retrieval, and utility of the tool are defined in performance measurement. In this study, the percentage of unanswered questions and the percentage of answered cases were determined for each item.
Data were analyzed using SPSS-24 software and p < 0.05 was considered a significance level.