Hypertension, often referred to as the “silent killer,” is a chronic health condition, affecting one in five adult Canadians,24
yet many people with hypertension are unaware of their condition and one-third of Canadians with hypertension have not achieved treatment targets.25
The HMI represents a multifaceted systems approach to treating hypertension in a primary care setting. The intervention’s effectiveness has been determined elsewhere.9
In this paper, we performed a cost-effectiveness analysis of this guidelines-based, chronic disease management model for primary care for patients diagnosed with hypertension. We found that the HMI was associated with reductions in the 10-year risks of CVD, ESRD, and an improvement in estimated life expectancy. This survival benefit was compared to the overall health care costs associated with the implementation of the intervention. Our analysis suggests that the HMI is a cost-effective means of providing primary care to patients with hypertension.
We predicted improvement in CVD, CHK, stroke, HF, PVD, and ESRD with the HSFO’s HMI. These results were robust to changes in age and imputation of missing data. Moreover, these improvements translated into a longer life expectancy.
Patients managed with the HMI had fewer predicted hospitalizations for CVD and ESRD, which translated into lower costs incurred by the health care system. We estimated the savings to the health care system for hospitalizations, diagnostic tests, physician services, and medications to be approximately $430 CAD per patient over their lifetime. The cost of the HMI itself was estimated at $29 CAD per 30 patient-days, based on total expenses incurred during this research study. Importantly, we suspect that this is an overestimate, as many of the included costs, such as consulting fees, IT development, and research-associated costs, are fixed and therefore become less significant when we extend the program to larger patient groups. If extended to the entire population, economies of scale would certainly result in lower costs per patient. We estimated that if the HMI costs $23.76 CAD per patient, this intervention would in fact be cost-neutral.
In our base case analysis, the HMI was cost-effective at an ICER of $4939 CAD for life-year gained, substantially below the traditional willingness-to-pay threshold of $50,000 CAD per life-year gained.26
This compares favorably with similar hypertension-targeted interventions, namely a nurse-administered telemedicine program (which delivered a series of educational and problem-solving modules to hypertension patients) at the Durham Veterans Affairs Medical Center primary care clinic (ICER $42,457 US to $87,300 US per life-year saved).28
Moreover, it compares favorably to other recently adopted cardiac technologies, such as an implantable cardioverter-defibrillator29
(ICER $213,543 CAD per quality-adjusted life-year gained) and drug-eluting stents30
(ICER $419,202 to $1,569,875 CAD per quality-adjusted life-year gained).
This intervention was targeted towards hypertension alone; however, as a benefit to more closely managed care, total cholesterol was also reduced. This phenomenon reinforces the potential applicability of this model to improving the management and outcomes of other chronic conditions and the importance of a comprehensive approach to primary and secondary prevention, targeting all risk factors.
This study must be interpreted within the context of several limitations. First, our estimates for the benefits of the HMI are based on 10-year risk from the Framingham population and extrapolations of life-tables from this dataset. These are not necessarily generalizable to real world effectiveness in Ontario. Second, we assumed that this reduction in the relative risk would translate directly to a decrease in hospitalizations for CVD and ESRD. The validity of these assumptions has not been empirically tested. Third, we found that the HMI led to a drop in blood pressure as well as in total cholesterol; however, given that the HMI is a behavioral intervention, we have no way to ascertain whether the effect of the intervention will persist for the rest of the patient’s life. In addition, our model did not consider quality of life differences between treatments. With more closely managed care, we would anticipate earlier identification of symptomatic deterioration and subsequent treatment. Therefore, we would expect that incorporating quality of life weights would amplify the differences observed between the HMI and standard care. Finally, our model was developed using multiple assumptions and extrapolations. As expected, there was a substantial amount of uncertainty in our estimates, as reflected in the wide confidence intervals around the ICER and in the CEAC curve, which showed that 39% of bootstrap replications were not cost-effective at a threshold of $50,000 CAD per life-year gained.
In conclusion, our analysis found that a community-based HMI is likely to result in meaningful health improvements at a reasonable cost. Further work would involve understanding whether the health changes associated with the HMI would be sustainable in the future. Policy implications of this work may involve widespread dissemination of the program within the province, and potential reductions in cardiovascular death and morbidity.