Cost-Utility of Triple Versus Dual Inhaler Therapy in Moderate to Severe Asthma

Jefferson Antonio Buendía; Diana Guerrero Patiño


BMC Pulm Med. 2021;21(398) 

In This Article


This study showed that triple was cost-effective than dual therapy in adolescent and adult patients with moderate-severe persistent asthma. Our findings support the GINA 2021 recommendations for using add-on LAMA to treatment with ICS-LABA asthma as an alternative in patients in Step 5. As we show this option generates 1.5 quality-adjusted life-year extra per patient concerning dual therapy with a cost of US$197, below of willingness to pay US 19 000 per QALY in Colombia. Using triple therapy emerges with our results as an alternative before using oral corticosteroids or biologics, especially in resource-limited settings.

Our results are in line with previous studies. Hyng et al., using a similar Markov model as our study, found, that in patients with poorly controlled asthma the adding tiotropium to ICS/LABA is a cost-effective alternative with an ICER $4,078/QALY in frequent SABA users and $8,332/QALY, on frequent exacerbators.[27] The differences in the magnitude of ICER are due to differences in the healthcare systems of Colombia and Korea and medical expenses. Indeed, our costs per event of OCR bust, ED visit, or hospitalization were 69%, 79%, and 46% less, respectively than in Korea. Wilson et al. using a six Markov model health states, estimate an incremental cost-effectiveness ratio of £21,906 per QALY gained being adding tiotropium to ICS/LABA cost-effective in the UK.[29] As is expected, the cost per event of an OCS bust, ED visit, or hospitalization was five times higher in the UK than in our study in Colombia; This can explain the differences in the magnitude of ICER between the studies. Zafari et al., using also a probabilistic Markov model with a 10-year time horizon and from a US societal perspective, found ICER of add-on therapy with tiotropium versus standard therapy, and omalizumab versus tiotropium was $34,478/QALY, and $593,643/QALY, respectively.[30] Despite differences in the model health states, higher costs of drugs and other direct costs in the US, and utilities, our conclusion is the same. One difference in our study to previous studies was the values of the utilities. The two previous studies used the utilities established in the Wilson study, which estimated them in the "PrimoTinAasthmatrial" population using the EuroQol EQ-5D tool in the UK population. We decided to use those reported in a systematic review to have broader values and in more diverse populations. Variations in the values of these utilities in the probabilistic sensitivity analysis did not significantly change the calculated ICER. Indeed, after of 10 000 simulations in our PSA tiotropium tends to be associated with lower costs and higher QALY; 80% of simulations were graphed in quadrant 1 of cost-effectiveness plane.

A not minor difference in our evaluation from previous studies is the fact that we have not only estimated the ranges of relative risks and transition probabilities using data from real-life studies but have adjusted our estimates for drug adherence. Assuming 100% adherence is unrealistic and tends to overestimate the effect of dual or triple therapy. A crucial methodological aspect is discussing willingness to pay (WTP) to declare Colombia a cost-effective technology or not. Since Colombia does not have a threshold that represents the WTP per unit of effectiveness (QALY), the ICER results per QALY were evaluated by using the reference corresponding to the World Health Organization (WHO) recommendation (three times the GDP per capita). Not having an own estimate of the WTP may be debatable; however, up to now, all the economic evaluations in health carried out in the country follow the threshold suggested by the WHO, which has also been endorsed by the national technology evaluation agency.[31] The results of the probabilistic sensitivity analyses confirm the robustness of the model results. Since relative risk and some transition probabilities and utilities do not come from the Colombian population, they were subjected to probabilistic sensitivity analysis as detailed below as recommended by Consolidated Health Economic Evaluation Reporting Standards (CHEERS) Statement.[20]

Our study has some limitations. We use utilities extracted from the literature and not estimated directly from our population. As was mentioned previously, the reliability and robustness of the results were evaluated by sensitivity analysis. Our results only refer to patients with severe asthma uncontrolled by medium-dosage to high-dose inhaled corticosteroids plus long-acting β2-agonists and cannot be extrapolated to patients with using oral daily corticosteroids. Studies of triple therapy have recruited both allergic and non-allergic asthma patients. By using evidence from such trials, we assumed the same health benefits of tiotropium for allergic and non-allergic asthma patients, and this assumption is supported by trials of tiotropium, which showed no difference between allergic versus non-allergic subjects.[7]

In conclusion, triple therapy in patients with moderate-severe asthma was cost-effective. Triple therapy emerges with our results as an alternative before using oral corticosteroids or biologics, especially in resource-limited settings.