Health worker knowledge of Integrated Disease Surveillance and Response standard case definitions: a cross-sectional survey at rural health facilities in Kenya

The first step in disease surveillance for frontline health workers is to diagnose and identify cases at health facilities. Rural health facilities in sub-Saharan Africa lack capacities to diagnose diseases through laboratory confirmations and health workers rely on clinical signs and symptoms to diagnose suspected cases of priority diseases. Validated clinical standard case definitions help standardize diagnosis of suspected cases across health workers within and across nations, and it is crucial for implementing the International Health Regulations 2005 [1‐3]. In addition, standard case definitions are important for managers at all levels to monitor epidemic trends across health facilities, investigate epidemic alerts, respond to potential outbreaks, and plan activities related to disease surveillance and control.

In Kenya as in other African countries, the World Health Organization regional office for Africa (WHO-Afro) implemented the Integrated Disease Surveillance and Response (IDSR) strategy [4, 5], which promotes the use of IDSR standard case definitions which are stipulated in technical guidelines, promoted through job aids, and included in the in-service training programs. In IDSR evaluation literature, processes of disease surveillance and activities supporting implementation have commonly been examined [6‐17]. However, the knowledge of IDSR standard case definitions by health workers and their supervisors have been assumed and no study has examined the awareness and knowledge of case definitions as stipulated in the guidelines [18, 19]. We therefore conducted a cross-sectional study among frontline health workers and their disease surveillance supervisors in rural Kenya to assess their knowledge of IDSR standard case definitions.

A cross-sectional survey was conducted as part of the mSOS (mobile Short-message-service based disease Outbreak alert Systems) trial [20], which measured the effectiveness of an intervention using text-message disease outbreak alert system. The study took place at rural health facilities and sub-county management offices in Busia and Kajiado counties in Kenya.

Correct knowledge of standard case definitions is an important first step for successful implementation of any disease surveillance strategy. Our study, one of the first that examined and quantified the knowledge of IDSR standard case definitions in sub-Saharan Africa, revealed alarmingly low levels of knowledge in Kenya. Patient registers at health facilities are used as source document for evaluating disease surveillance reporting practices and health systems in general [35], and our study findings imply that the validity of the source documents may be overlooked. First, we found that knowledge of dengue is practically non-existent among health workers and their disease surveillance supervisors. We speculate that the relatively complex case definitions, lack of routine laboratory diagnosis to confirm dengue in rural health facilities, similarities with other mosquito-borne diseases such as malaria, and the possibility of perceived non-lethal nature of the disease may contribute to low knowledge. The findings are however not unique to our study sites since suboptimal knowledge of dengue has also been reported among health workers in high-income countries [36]. Similarly to dengue but even more surprising finding was found with respect to measles. Despite simple case definitions, only a quarter of supervisors and just one in ten health workers could state five clinical criteria, which have been promoted in Kenya over a decade. It should also be acknowledged that surveillance guidelines in Kenya include an ambiguous and unnecessary wording, “any person in whom a clinician suspects measles”, which undermines the validity of promoted case definitions and itself precludes standardization. Similar sentiments were expressed in an evaluation conducted in South Sudan [37]. The most unexpected finding was observed on dysentery where only half of the health workers were able to correctly state the simple criteria of “bloody diarrhea.” Health workers responses of other dysentery related symptoms such as “abdominal cramps and pains” may suggest the health workers are knowledgeable of the text book clinical features [38], but disregarding the IDSR guidelines.

Our predictor analysis examining factors for correct knowledge added further light on what may influence knowledge of dysentery. First, we found that nurses are significantly more knowledgeable compared to other cadres of health workers. One of the possible explanations is that health workers with greater pre-service clinical training are more likely to overrule guidelines, which has been shown in previous studies [39, 40]. Second, the presence of a simple job aid in the form of a case definition poster at a health facility was significantly associated with better knowledge. Reminders are shown to be effective intervention in other studies [41], and policy makers should prioritize such seemingly simple intervention. In addition, this study was conducted in the context of evaluating a text message intervention to enhance disease surveillance reporting [20] and reminder interventions through such channels but targeting health workers should also be explored in the future. Finally, exposure to either refresher training or supervision was not associated with greater knowledge of case definition. We do not think that these interventions are ineffective per se, but perhaps the sub-optimal quality of the training and supervision intervention may have contributed to the analysis results. The study did not have a component to measure the quality of training or resources to follow-up supervision mechanisms after training, but we speculate that these factors may have contributed to low knowledge levels. Moreover, the low knowledge levels of disease surveillance supervisors found in this study further elevates the levels of challenges faced in the field to promote adherence to IDSR standard case definitions.

Several study limitations should be acknowledged. First, the knowledge levels were self-reported by health workers and the study did not examine whether the health workers’ knowledge was translated into practice. We also did not examine the duration of professional experience that may have influenced the level of knowledge. Knowledge is however a prerequisite for correct practice and it is unlikely that health workers would adhere to guidelines. We also cannot confidently assume uniformity in diagnosis of suspected cases when knowledge of the standard case definition is not present. Second, we limited the analysis to three tracer diseases and applied a rather relaxed approach to interpreting the definitions. Given the low knowledge levels found overall in the study, stricter case definitions would have however yielded even lower knowledge results, and likely without changes to the interpretation of the findings. Third, the small sample size used in this study may have precluded showing significant associations when associations were truly present, and results should be interpreted with caution. Finally, while the study included a number of health workers from two areas in Kenya, the results are not nationally representative and cannot be generalized across the country.

Overall, the IDSR case definition knowledge needs to be further improved among both health workers and their supervisors, and the findings are likely to impact the reliability of routine surveillance reports generated from health facilities. The findings also suggest that simple intervention such as reminders are more effective than ad hoc trainings and routine supervision to ensure uniformed knowledge of case definitions. Further studies on examining effective methods of enhancing knowledge of IDSR standard case definitions are justified.