[ad_1]
Sociographic data
The 15 health decision-makers interviewed were aged between 29 and 60 years. Most participants were below the age of 40 years. Respondents consisted of 14 male and one female health decision-makers. All held university degrees: Bachelor (n = 1), Master (n = 6), MD (n = 1), MD and Master (n = 4), PhD (n = 3). They were working for the Federal Ministry of Health (FMOH) (n = 5), the Ethiopian Public Health Institute (EPHI) (n = 3) and other Ethiopian institutions like the Ethiopian Biotechnology Institute (EBTI) (n = 1), Black Lions hospital (n = 1), St. Paul’s hospital (n = 3), Family Guidance Association of Ethiopia (FGAE) (n = 1); one was a public health officer (n = 1). Participants engaged in high-level and low-level decision-making, from FMOH to field workers (n = 2). An overview of the sociodemographic data is provided in Table 1.
The following six sub-themes were identified as important considerations for participants in the context of achieving greater data reliability and dealing with resource constraints: data resources; HPV and CC burden in Ethiopia, HPV screening and vaccination, information for decision-making, significance of mathematical modeling and computer simulations and outlook and recommendations of decision-makers. The themes will be presented in this order.
Data resources
The theme ‘data resources’ covers the preferred information resources of our participants. Health decision-makers drew on national and international data sources for their work and every decision-maker we talked to spoke of data-scarcity (Box 1, excerpt 1).
In general, decision-makers used international data from the World Health Organization (WHO), particularly from the GLOBOCAN database kept by the International Agency for Research on Cancer (IARC) for planning and writing proposals. Additionally, they said that they rely on international publications from the American College of Obstetricians and Gynecologists (ACOG) and International Gynecologic Cancer Society (IGCS), as well as international conferences on CC and human papillomavirus (HPV) research. One-third of the health-decision makers expressed doubts about the international data sources (Box 1, excerpt 2).
Local data sources were considered more useful by almost every participant in our study. A frequently named source was the District Health Information System 2 (DHIS2), which gathers local data and is run by the Federal Ministry of Health (FMOH) (Box 1, excerpt 3). The Demographic & Health Survey (DHS) implemented by the Central Statistical Agency (CSA) of Ethiopia in collaboration with the FMOH and Ethiopian Public Health Institute (EPHI) is a survey of representative samples on the national and regional level in Ethiopia28. The DHS was frequently mentioned by health decision-makers (Box 1, excerpt 4) and other local sources of HPV and CC data specified were a cancer registry in Addis Ababa as an individual effort of local researchers. Moreover, interview participants also highlighted that there are useful local studies from hospitals identifying genotype distribution of HPV infections and cervical cytology abnormalities like Ali and colleagues;29 Addis Ababa University, school of public health; community-based surveys and HPV prevalence studies conducted by students; the HPV consortium, which is a collaboration of different public institutions of Ethiopia situated at the Ethiopian Biotechnology Institute (EBTI) in Addis Ababa.
HPV and cervical cancer burden in Ethiopia
With an estimated 6300 new cases in 2018, health decision-makers regarded CC as the second most frequent malignancy among women in Ethiopia. Economic and gender vulnerabilities were mentioned by interviewees as barriers to utilizing healthcare services and spoke of the issue of patients presenting themselves in a very late stage of the disease to their healthcare providers. According to participants, reasons for seeking healthcare services late might be the fear of not being able to pay for the diagnosis or treatment, being unaware that the service is free or being denied permission by the husband. As one of our participants put it (Supplementary Table 1, excerpt 1):
These barriers towards a quick diagnostic might lead to undetected CC cases and deaths. which are not reflected in the official data – “because there are people dying in their home…”. Undiagnosed affected women were thus unable to get help.
The HPV burden in Ethiopia and its geographic focus remained unclear in interviewee’s descriptions. Interviewees hypothesized that urban areas were more likely to be affected than rural areas due to the spread of sexually transmitted diseases (Supplementary Table 1, excerpt 2).
Furthermore, HPV prevalence goes hand in hand with HIV prevalence, which was perceived to be higher in the cities (Supplementary Table 1, excerpt 3).
Ethiopian cross-border traffic was seen as another aggravating factor (Supplementary Table 1, excerpt 4). It was local health decision-makers, mainly male field workers, who provided this information.
HPV screening and vaccination
All participants found HPV screening and vaccination to be a priority and depending on their expertise, decision-makers were more inclined either towards screening or vaccination as a priority. In regards to vaccination, prevalent genotypes of HPV were considered crucial. The general approach relies on genotypes 16 and 18 as the most predominant types and all decision-makers underscored the importance of HPV-16, but there were doubts upon HPV-18. This is because even though HPV-18 is considered a critical trigger of CC in Ethiopia, health decision-makers had their doubts regarding ranking it second after HPV-16. Interviewees with higher professional positions casted doubts upon HPV-18 because they felt, based on isolated research, there might be genotypes that are more common than HPV-18 (Supplementary Table 2, excerpts 1-2).
The health decision-maker and interviewer then deliberated the significance of HPV type 16 and type 18 for the Ethiopian society and their relevance for incorporating particular kinds into a mathematical model in the following conversation. It appeared that the interview partner changed their mind in this deliberation, as the two interview partners agreed on the necessity of further studies to verify the relative importance of genotype HPV 18. Both parties agreed on focusing on the most prevalent types for integration into a mathematical model (Supplementary Table 2, excerpt 3).
The sharing of the interviewer’s personal experience in a local Ethiopian hospital revealed the interviewee’s position on screening and prevention. The interviewer explained the encounter with a patient suffering from advanced CC without hope for a cure (Supplementary Table 2, excerpt 4).
A joint takeaway for both interview partners was the affirmation of the relative importance of screening and prevention programs on CC.
Furthermore, due to the conversational nature of the interview, a range of religious and cultural variables seemed to present as impacting whether or not a young Ethiopian girl is vaccinated against HPV (Supplementary Table 2, excerpt 5).
As the interview partners stated scarce resources are a concern not only in research but nearly everywhere across the healthcare system. Interviewees perceived a general lack of health professionals (Supplementary Table 2, excerpt 6), and treatment and screening for CC was depicted as limited to big cities leaving the rural areas even more resource-deprived. Long waiting lists for surgery and radiotherapy were discussed with participants describing that Ethiopia currently has only one radiotherapy facility for advanced CC treatment. Furthermore, the total health expenditure in the country was seen as low and particularly low for cancer prevention and treatment. Moreover, they noted that many health initiatives, such as CC preventive programs, are donor-dependent and that the scarcity of resources was regularly leading to ethical dilemmas (Supplementary Table 2, excerpt 7).
Information for decision-making
To make an informed decision about HPV interventions and CC prevention, health decision-makers discussed the necessity for specific types of information. It emerged that the specific burden of CC, as well as HPV infections, is crucial for health professionals to know on a national and regional level. According to interviewees, researchers have to identify the seroprevalence of HPV on a national level and in local populations and that the information about HPV was required to be as detailed as possible (Box 2, excerpt 1).
Decision-makers would like to have exact information about herd immunity, as well as HPV hot spots which were defined as a town or region with a high prevalence of HPV. Furthermore, they expressed the need for information on the potential risk factors associated with HPV.
The availability of resources, as well as the identification of resource-poor areas, were deemed essential by decision-makers. Above all, it emerged from interviews that health decision-makers were interested in the cost and sustainability of intervention and if the intervention was cost-effective and would have a continuous positive effect on the population (Box 2, excerpt 2). Vaccination thresholds of herd immunity were also of interest to all participants but one. One decision-maker proposed that there should be a vaccination for every girl in the country.
Participants said that the factors that determine the uptake of vaccination or acceptance of health interventions concerning HPV and CC were of importance because they were also worried about who is visiting the health facilities and who is normally interested in receiving assistance versus who is not.
Participants asked that information presented to stakeholders using the predictions be of good quality and backed by reliable data (Box 2, excerpt 3). Results should be displayed clearly and understandably.
Significance of mathematical modeling and computer simulations
Knowledge of mathematical modeling and computer simulations differed among participants with some being experts in statistical modeling or having at least some experience while others did not have any experience. The majority of health decision-makers found mathematical models and computer simulations useful for their decision-making and would use the information generated in their future work or were using it already. They considered information from these models useful as long as the quality of input data was good (Supplementary Table 3, excerpt 1), and the model had been peer-reviewed. Younger participants argued more enthusiastically in favor of theoretical predictions with computer simulations. Seven of nine participants aged 29–40, and one of six participants aged 41–60, raised fewer concerns towards these predictions. Furthermore, in the former age group, 3/9 and in the latter, 3/6 were less experienced with modeling topics.
As discussed, and agreed upon by the interviewer and interview partners, there was a trade-off in HPV spread modeling between a simple model with a small confidence interval of its predictions, which models only a small part of reality with greater precision, and a sophisticated model with a larger confidence interval of its predictions, which models a more realistic environment with less precision. A conversation about simple and sophisticated mathematical models helped to contextualize their usages (Supplementary Table 3, excerpt 2).
Both interviewer and interviewee preferred a simple model but agreed that the Ethiopian context had to be considered when choosing an appropriate model and that users of a model must acknowledge the trade-off between a simple and a sophisticated model.
Most participants said that they preferred a simple model over a sophisticated one, preferring less realistic but more precise predictions over high-resolution models. Some decision-makers would like to start with a simple model and move towards more sophisticated ones over time. Health decision-makers generally trusted mathematical models and having a team of specialists was viewed as helpful to solve the trust problem concerning models and computer simulations, as generated from this piece of conversation (Supplementary Table 3, excerpt 3).
One participant expressed the wish to understand the mathematical model more deeply to trust it (Supplementary Table 3, excerpt 4).
Outlook and recommendations of decision-makers
Mathematical modeling and computer simulations were seen favorably by health decision-makers and were either being used by decision-makers, or they were very interested in adopting them in the future. They hoped predictions will help with various problems in the country and encourage to promote interventions. As interviewees suggested, models might aid in determining the incidence and prevalence of CC, as well as the most common serotypes. A decision-maker expressed optimism about the development of an HPV transmission model and emphasized the importance of the input data quality (Box 3, excerpt 1).
Digitization and computer modeling is perceived as a chance (Box 3, excerpt 2).
However, health decision-makers had a few recommendations for modelers working with mathematical models. Obtaining accurate baseline data and emphasizing high-quality input data was highlighted as most important stage for generating credible predictions. One participant just exclaimed: “Data, reliable data!”(….) when asked for recommendations.
Furthermore, because Ethiopia has many ethnic groups, each with their own culture, religion, and language, health decision-makers believed that taking cultural factors into account in the local Ethiopian setting is critical (Box 3, excerpt 3).
Participants emphasized the necessity of re-evaluating and discussing a mathematical model after introducing it and underscored the need for constant improvement (Box 3, excerpt 4). Specific recommendations were: consider immune profiles; model investment return; talk to the fieldworkers; try to promote the importance of modeling; use software which is freely available in Ethiopia and present the model in simplified ways.
[ad_2]
Source link
