Approved by an Institutional Ethics Committee and

¶ … approved by an institutional ethics committee and designed to conform to the Helsinki Declaration (Zaman, Sheikh, Das, Zaman, & Pal, 2014). Informed consent was obtained after explaining that participant confidentiality would be protected. The risks were minimized by having the participants complete the questionnaires at home and the benefits maximized when visited by clinician/researchers during data collection.

The purpose of the study was to identify demographic variables predicting delays between symptom onset and TB treatment-seeking behavior, so the gathering of demographic information through self-reports was appropriate and rigorous, but not the most rigorous. A more rigorous approach would be to validate collected information using family member interviews; therefore, key variables were not operationalized using the best possible method. Given the study design and purpose, a comparison group was not needed.

Since the researchers were gathering demographic and symptom onset information, the risks to internal and construct validity was minimized. Socioeconomic variables were then identified that would allow clinicians to predict which groups were most likely to delay treatment following symptom onset, thereby maximizing external validity and providing congruency between the operational and conceptual definitions. Since extensive demographic information was gathered, the population was described extensively (Zaman et al., 2014).

Limiting participants to all smear positive patients identified in Assam, India during 2007 provided the best sampling strategy, maximized representativeness, and minimized sampling bias. Accordingly, sample size, statistical power, and sample bias considerations were irrelevant. The instrument was not described at all, except to declare that it was designed by the researchers with the help of experts. The data obtained was consistent with the findings of other published studies, thereby conferring validity and reliability to the findings.

Qualitative: Methods The National Tuberculosis Control Program in Burundi approved of the study and nurses who agreed to be interviewed did so voluntarily and with informed consent; therefore, the recruitment approach was appropriate and ethical (Carlsson, Johansson, Eale, & Kaboru, 2014). Interview transcripts were coded and all recordings destroyed, so the risks were minimized; although, benefits were not obvious. The study design was descriptive and used qualitative semi-structured interviews, which was congruent. The researchers and trained interpreter travelled to clinics to conduct the interviews, which took between 15 and 40 minutes.

Sufficient time was allowed in the field to collect data and saturation had been reached by the seventh interview, suggesting the best possible method of sampling had been used given the purpose of the study. Eight nurses at two clinics were interviewed and the demographic and setting information was adequate, but only one contact was made and triangulation was not used. While a lot of data was gathered, at least one subcategory provided only minimal information.

This suggests insufficient data was gathered to address each issue comprehensively in the study, thus limiting depth and richness. The two interviewers analyzed the data together using content analysis, an iterative process that repeatedly reviews the transcripts and identifies subcategories and categories relevant to the study purpose. A third co-author read through the data to confirm it supported the categories, thereby increasing its trustworthiness (validity). This was the limit of what the authors did to ensure trustworthiness, so more could have been.