This study has some limitations. First, the data set came from a single randomized control trial that may not reflect outcomes in general practice. Moreover, the authors of this trial noted that the ICU LOS “should be interpreted with caution due to the low numbers.” Results of the model may also present some variability from one country to another, especially regarding hospital practice. For example, for payers that pay institutions on a per diem rate, and adjust for ICU admissions, the ICU cost-savings evidenced in this study as a driver to justify the PISC disappear. It is not clear whether there is a true difference in ORTHO LOS between the intervention and control groups in reality (the difference observed in the study by Møller et al3 was not significant), and what would be the impact of such a difference. Our study indicates that such an impact would be negligible in France (see scenario A), but estimations of ORTHO LOS would benefit from a greater precision because differences in ORTHO LOS in the intervention and control groups may become a substantial element of the costs in some systems favoring a per diem rate. The results were not sensitive to the diagnosis-related group of patients (Table 2, scenarios G and H) and were only moderately sensitive to the daily cost of intensive care (scenarios B and C). The absence of available data on such a topic requires future research with large samples to estimate more precisely the costs of the complications when patients are transferred to ICUs. The results were more sensitive to the probabilities of complications in the intervention and control groups (Table 2, scenarios D, E and F), potentially reflecting the success of PISC in terms of smoking cessation. In the study by Møller et al,3 the rates of smoking cessation were very different in the intervention and control groups (64% and 7.7%, respectively). Even if a PISC is likely to be applied only to highly motivated patients, our results indicate that such an intervention must be both inexpensive and successful if considering the short-term benefits. Nevertheless, our sensitivity analyses indicate that a neutral cost-benefit ratio would require a rate of complications in the intervention group 300% higher than reported in the literature (Table 2, scenario E), or a mean ICU LOS of 3.16 days (close to the value of 3 used in scenario F), 63% lower than reported in the literature. Still, the presented results are limited to total hip and knee arthroplasty patients and, therefore, should not be considered as valid for other types of procedures for which the impact of a PISC remains to be estimated. Another limitation is the interval between surgery and smoking cessation. Møller19 reported that for most patients, the interval between the initial consultation and surgery was shorter than the 6 to 8 weeks on which the initial study was based. In their review, Theadom and Cropley5 indicate that only two studies specified the exact period of smoking cessation, and that no “optimal” period of preoperative smoking cessation could be identified from the available evidence. Nevertheless, our study, based on the only available data compatible with a cost-benefit analysis, indicates that PISC is highly beneficial for total hip and knee arthroplasty patients. The modest cost of such interventions is more than offset by short-term savings in hospitalization costs, owing to a reduction in postoperative complications. Despite the limitations mentioned, the present study also provides an approach to modeling cost and benefit of PISCs that could be adapted to other settings and other patient conditions.