|Home | About | Journals | Submit | Contact Us | Français|
A cost effective use of public health resources
Barham and colleagues1 provide a valuable review of cost‐effectiveness studies of one to one sexually transmitted infection (STI) prevention interventions; one to one interventions are those delivered to one person at a time, such as screening or individual counselling (as opposed to group counselling, community‐level interventions, and so on). Their review summarises the findings of 55 cost‐effectiveness studies, most of which focus on chlamydia and HIV prevention interventions. A more comprehensive version of their review providing methodological details and a listing of hundreds of other cost‐effectiveness studies excluded from their review is available online.
One to one STI prevention interventions were found to be cost effective in a vast majority of the studies in the review. This abundance of evidence for the cost effectiveness of one to one interventions adds significant support to the idea that risk reduction counselling should be part of the standard of STI care.2,3
Barham and colleagues1 rate the quality of each study included in the review. These ratings should be viewed with caution, owing to the difficulties in developing and applying an appropriate set of criteria. The quality ratings of the dynamic and static models (those that incorporate population‐level transmission dynamics and those that do not, respectively) may reflect a preconceived notion that cost‐effectiveness studies based on dynamic models are automatically of higher quality than those based on static models. For example, all of the studies based on dynamic models scored the highest possible quality rating compared to less than 10% of the studies based solely on static models. Although dynamic models are needed to advance our understanding of the costs and benefits of STI prevention strategies, we need not dismiss static models in this pursuit. The incorporation of STI transmission dynamics is inherently uncertain, and dynamic models are not necessarily more accurate than static models. As George Box reportedly stated, “All models are wrong, but some are useful”.4 Static and dynamic models have been and should continue to be useful tools in the field of STI prevention.
Cost‐effectiveness studies, like all other studies, have limitations. For example, as Barham and colleagues1 note, the existing literature on the cost effectiveness of chlamydia screening has been questioned because of the reliance on static models and the uncertainty in the probability of developing pelvic inflammatory disease.5
However, perhaps the main reason why a study's estimate of an intervention's cost‐effectiveness ratio differs from the intervention's “true” cost‐effectiveness ratio is not a shortcoming of the model applied or the choice of a particular parameter value, but rather the omission of a wide range of benefits of STI prevention—such as the impact of STI prevention on HIV incidence, the indirect costs (such as lost productivity) and intangible costs (such as pain and suffering) averted by STI prevention, the impact of the intervention on other STIs not addressed in the analysis, and a host of other benefits, which, if included in the analysis, would result in far more favourable cost‐effectiveness estimates. The prevention of STI‐attributable HIV cases can be a valuable outcome of STI prevention activities.6 Because the average lifetime direct medical cost per case of HIV is roughly 1000 times that of gonorrhoea or chlamydia,7,8 even miniscule impacts of an STI prevention intervention on STI‐attributable HIV cases could have a sizeable impact on the estimated cost effectiveness of the intervention. The indirect costs of STIs can be substantial as well. A recent study suggested productivity losses of about $130 per case of untreated chlamydia in women:9 roughly half the estimated lifetime direct medical cost per case of chlamydia in women.8 Similarly, the inclusion of intangible costs would also lead to more favourable cost‐effectiveness estimates. Although the intangible costs of STIs are virtually impossible to assess, the intangible cost burden could easily outweigh the direct medical cost burden.
One way to incorporate the indirect and intangible costs averted by an intervention, at least in part, would be to assess the intervention's impact on quality of life, using a metric such as quality‐adjusted life years (QALYs). As noted by Barham and colleagues,1 very few of the studies in their review applied QALYs. One of the exceptions was a study that found that chlamydia screening would cost $2000 to $7500 per QALY saved (in 2000 US dollars),10 which is quite favourable compared to many other public health interventions. The usefulness of future cost‐effectiveness studies could be greatly enhanced by the use of QALYs, and the call by Barham and colleagues1 for the development and refinement of STI‐specific QALY estimates is certainly reasonable.
The estimation of the cost effectiveness of STI prevention programmes is not an exact science. However, it is known that STIs impose a considerable health and economic burden, and that one to one STI prevention interventions are effective in preventing STIs. As such, it is no surprise that the vast majority of studies reviewed by Barham and colleagues1 find that one to one STI prevention programmes are a cost effective use of public health resources.
Competing interests: None.
The findings and conclusions in this commentary are those of the author and do not necessarily represent the views of the Centers for Disease Control and Prevention.