According to the CDC Guidelines, “predictive value positive (PVP) is defined as is the proportion of reported cases that actually have the health-related event under surveillance”. Mathematically, PVP is calculated as:
True Positive + False Positive
The assessment of sensitivity and PVP provide different perspectives regarding how well the surveillance system is operating. The PVP reflects the number of people without trachoma incorrectly included in the surveillance system. As already discussed, eye health experts consulted advised that it is highly unlikely that appropriately trained screening staff will identify trachoma in a child that does not have trachoma (false positives). Therefore, it is reasonable to conclude that the likelihood of false positives in the national trachoma surveillance data is low and consequently PVP is high. Calculating PVP empirically for the national trachoma surveillance system would be a major exercise. An external measurement (‘gold standard’) of whether a child diagnosed with trachoma actually had trachoma would be required (the process would probably involve re-screening by an expert practitioner).
Given that the number of false positives in the national trachoma surveillance system is likely to be low (and hence PVP is likely to be high) there is little value in mounting an exercise to make a ‘gold standard’ measurement. The concern with PVP arises from the fact that false positive reports can lead to unnecessary interventions, and falsely detected outbreaks can lead to costly investigations and undue concern in the population under surveillance88. This problem is not serious in the context of the national trachoma surveillance system, as the intervention is not risky or costly, simply the administration of antibiotics and the provision of preventative health information. Hence diverting resources from screening and intervention to accurately measuring PVP is not considered valuable for the trachoma surveillance system.
Findings: With respect to the predictive value positive of the national surveillance system the evaluation team finds:
- the current system has high PVP, as the risk of false positives (diagnosing a child with trachoma who does not have it) is small;
- there is little value in pursuing an accurate measurement of PVP as the resources used to treat identified cases are not costly and the risks are not high.top of page
88CDC (2001) Updated Guidelines for Evaluating Public Health Surveillance Systems. Recommendations and Reports July 27, 2001/50(RR13);1-35