2.1. iPCQ and Other Instruments
The iPCQ has five domains with 18 items. Among these, the domain of general questions regarding the target population (Item A1–Item A6) is used to estimate the characteristics of a patient. The other domains (including general questions regarding paid work [Item 1–Item 3], absenteeism module [Item 4–Item 6], presenteeism module [Item 7–Item 9], and unpaid work module [Item 10–Item 12]) are required to identify and value the loss in productivity [13
The general questions regarding the target population identify the respondent’s characteristics including survey date, age, sex, level of education, and type of work. Researchers can perform subgroup analyses based on this domain. The general questions regarding paid work survey the characteristics of paid work, including work periods (in hours and in days) per week. Based on the answers, researchers can calculate the average work hours per day. The domains of each module (absenteeism, presenteeism, and unpaid work) comprise items that measure the characteristics of productivity loss (whether productivity loss occurs and the period of productivity loss for each domain). In addition, each module has distinct questions as follows: in the absenteeism module, there is a question to identify whether the absenteeism is short-term or long-term (Item 5). In the presenteeism module, Item 9 asks how much the productivity efficiency of paid work has been reduced compared to that under normal conditions, by measuring the extent of health-related problems on a scale of 0 to 10 points. In the module for unpaid work, item 12 measures how health-related problems have reduced the productivity of unpaid work, based on a “third-person criterion.” [7
2.2. Study Design and Process
We conducted the translation and cross-cultural adaptation of the iPCQ into Korean (including forward translation, back-translation, pilot test and cognitive debriefing, and finalization). Next, we analyzed the validity and reliability of the Korean version of iPCQ, including feasibility, concurrent validity, construct validity, and test–retest reliability.
To perform the validation study, we conducted surveys including the WPAI and the Short Form 36-Item health survey (SF-36), as well as the iPCQ. The WPAI is the most frequently used instrument for measuring productivity loss in Korea. This instrument comprises six items (currently employed [Item 1], hours missed due to health problems [Item 2], hours missed because of other reasons [Item 3], hours actually worked [Item 4], degree to which health affected productivity while working [Item 5], and degree to which health affected unpaid activities [Item 6]) and can estimate four main scores (including percent work time missed due to health, percent impairment while working due to health, percent overall work impairment due to health, and percent unpaid work impairment due to health) [15
]. The SF-36 is an instrument designed to identify a patient’s quality of life and has been culturally adapted into Korean and validated [16
]. This instrument includes 36 items creating 8 scales (physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health) [17
2.4. Translation and Cultural Adaptation
We developed the Korean version of the iPCQ based on the guidelines for cross-cultural adaptation, including the process of preparation, forward translation, back translation, pilot test, cognitive debriefing, and finalization [19
]. In the preparation step, we contacted the iPCQ development team (Institute for Medical Technology Assessment, iMTA) to grant permission for development of the Korean version of iPCQ through the iMTA website (https://www.imta.nl/questionnaires/
). We explained the goal of our research and source of funding to the development team. We acquired the approval in May 2017.
In the forward translation, two independent translators, who were native Koreans with a good command of English, conducted forward translations of the original iPCQ into Korean. Our research team (H.K., S.H., S.K., and H.S.S.) reconciled the two forward translated Korean versions of the iPCQ into a single Korean version considering conceptual equivalence and cultural appropriateness.
In the back translation, two translators who are bilingual in English and Korean, without knowledge of the original iPCQ, received the reconciled forward-translated Korean version and independently conducted back translations. Our research team (H.K. and H.S.S.) reviewed and compared the conceptual equivalence and cultural appropriateness between the original version and the back translations. We repeated the discussion process until a consensus on the translation had been achieved, and then prepared the intermediate Korean version of the iPCQ.
We conducted a pilot test using the intermediate Korean version of the iPCQ with 15 outpatients (≥19 years) visiting gynecology clinics (7 patients in the Korean Medicine Hospital of Daejeon University, and 8 patients in the KyungHee University Korean Medicine Hospital). In the pilot test, we distributed the intermediate Korean version of the iPCQ to these patients and conducted cognitive debriefing. At the end of the questionnaire, we asked the patients the following questions: 1. “Was this question difficult to answer?”; 2. “Was this question confusing?”; 3. “Was this question difficult to understand?”; 4. “Was this question upsetting or offensive?”.
For finalization, a research member (H.K.) prepared a record sheet including the process of the five translations (two forward-translated versions, one reconciled version, and two back-translated versions) and the results of the cognitive debriefing. The research team (H.K. and H.S.S.) checked all issues and changes from the original to the translations and patients’ comments. We repeatedly reviewed and discussed the cultural appropriateness, conceptual equivalence, and difficulty of the translations until arriving at a consensus. Then, we developed the final Korean version of the iPCQ. We sent the final translations to the iMTA development team for final approval. After minor modifications of the instrument name and acknowledgement, we received approval for the Korean version of iPCQ.
2.5. Validation and Statistical Analysis
We conducted the validation study with outpatients (≥19 years old) who visited a gynecology clinic from August to September 2018. Before patients received medical care, we conducted surveys on the socio-economic characteristics, iPCQ, SF-36, and WPAI. We estimated the feasibility, including the mean respondent time spent filling out the questionnaire and the proportion of missing values for the Korean version of iPCQ [14
]. We administered the iPCQ again after patients received medical care for the test–retest reliability.
We used the test–retest reliability measure to identify the extent to which the scores of the Korean version of iPCQ are similar at different times (2-h gap between before and after receiving medical care). We used the intra-class correlation coefficient (ICC) considering the following cut-off values: <0.5 (poor), ≥0.5 (moderate), ≥0.75 (good), ≥0.9 (excellent) [21
We conducted the concurrent validity test to analyze the extent to which the scores of the Korean version of iPCQ correlate with an external criterion [22
]. We used the Korean version of the WPAI:GH, which is a validated and widely used instrument in Korea, as the external criterion [15
]. First, we divided the measurement domains into absenteeism, presenteeism, and unpaid work by referring to the manual for each instrument of the iPCQ and WPAI. Next, we measured the time related to productivity loss in each domain based on the results of the survey. Finally, we conducted a correlation analysis for each domain (absenteeism, presenteeism, and unpaid work) between the two instruments. Because all the variables we compared were not normally distributed, we used the Spearman’s rank correlation coefficient ρ (|ρ| < 0.3: low, 0.3 ≤ |ρ| < 0.6: moderate, 0.6 ≤ |ρ|: high) [25
We performed the construct validity to evaluate whether the productivity loss measured by the Korean version of iPCQ is correlated with the SF-36 domains (including physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional, mental health) using the Spearman’s rank correlation coefficient ρ. We hypothesized that each domain of productivity loss (absenteeism, presenteeism, and unpaid work) estimated by the iPCQ was negatively correlated with the quality of life. In general, it is known that a patient’s health problem has an effect on the measurement of the health-related productivity loss [25
]. We classified the correlations into the follow categories: high correlation (r ≥ |0.3|) and low correlation (r < |0.3|) [27
]. We used STATA version 15.0 (StataCorp, LP, College Station, TX, USA) for calculating the ICC. For the other analyses, we used SAS version 9.4 (SAS Institute, Cary, NC, USA).