Comments on Interpretation of Findings in Article on Primary Health Care in Rural Malawi Nadi Kaonga, Columbia University 24 April 2013 Dear Editors of the BMC Health Services Research Journal and Authors of the Manuscript by Makaula et al (2012), I hope this finds you doing well. I am writing in regards to the research article, "Primary health care in rural Malawi - a qualitative assessment exploring the relevance of the community-directed interventions approach" by Makaula et al (2012) [http://www.biomedcentral.com/1472-6963/12/328]. There are several times throughout the article where the interpretations of the findings seem rather liberal and there is the question of consent. I have outlined a few of those issues below. I hope that these will be taken into consideration. [Please feel free to pass along my comments to the appropriate persons.] 1. Selection of the Study Districts The authors note that the study districts were selected for their similarities. However, when one looks at Table 1, it is clear that on almost all points, Mzimba District has outperformed Mangochi District. [This data follows national trends of the Northern Region having better health and education indicators than the Central and Southern Regions, as observed over the years in Demographic and Health Survey compilations.] The only arguably similar indicators of the districts were the following: "number of health facilities per 100,000 population", "number of [...] CHWs per 100,000 population", "percent of the population with safe water" and "GDP% of population under the poverty line". If the authors were looking to have more similar districts, then other selections should have been made. However, if they were hoping to show that districts in different parts of the country had a similar outlook for primary health care despite their comparative difference on key indicators, then this should have been the point emphasized by the authors. 2. Ranking of Topics by Respondents The authors claim that the respondents had brought up and ranked similar issues. However, tables 4 and 5 call for a different interpretation of the data. While there was some overlap in priority health issues, the community primarily was concerned with infrastructural issues rather than disease, and the providers were primarily concerned with diseases (table 4). Despite that difference in the overall list, malaria was a top concern across the two groups (table 4). While there was at least some minimal overlap across groups when the participants were stratified by gender and age, we see that the different age groups had different priorities (table 5). The priority lists of each group were more different than similar, especially if we are assuming that the frequency that such issues were mentioned matters. [This may be due to their stage in life and roles in the household. It would have been more informative if the authors could share information on that based on their data, if possible.] 3. Oral Consent The authors noted that they obtained oral consent from the respondents. I am curious as to why only oral consent was obtained across all the participants. Some of the respondents (i.e., health care providers) most likely could have provided informed written consent. 4. Lack of Observation Component While the authors had taken the time to review district-level documents, to make the claims that a functional health system is in place, an observational component would have further supported (or disputed) their claim. For example, one may see information in the documents about bed net distribution, but that is not enough information. [Nor does the provision of bed nets alone constitute as primary health care.] It is important to also see how the bed nets are being used/if the community has adopted the use of the bed nets all in relation to malaria health outcomes. I would imagine that a functional health care system would not only have a means to distribute the bed nets, but would also have systems put in place to encourage their proper use. Sometimes, such information can only be gleaned through observation and discussion and not through documents alone. 5. Functional Health System I would disagree with the authors' results that "the findings show that there is a functional [Primary Health Care] system in place in the two study districts...". Rather, the data indicate that there is promise of a functional system as the framework seems to be in place, but the infrastructure, resources and community investment pieces still need to be addressed. In conclusion, I hope that my points above have been clear and that they underline some of the limitations of how the data has been interpreted in this article. I would like to note that, as a Malawian and public health practitioner, I was thrilled to see such research efforts taking place, and hope that my feedback will only strengthen the research and help move forward the agenda of improving Malawi's health care system. Thank you for your time. Competing interests I have no competing interests to declare.