Model

An individual-level (microsimulation) model was developed, representing hypothetical individuals with confirmed LTBI who were initiating drug therapy in Canada (figure 1). Ethics approval was not required for this simulation-based study. Individuals entered the simulation in a decision tree that represents the first year of LTBI drug therapy. Based on the individual’s adherence to drug therapy and the drug regimen completed, simulated individuals entered a Markov model with a differential lifetime reactivation risk for TB. TB reactivation could occur any time in an individual lifetime, and individuals were assumed to receive TB drug treatment. On successful completion of active TB therapy, individuals were at heightened risk of TB relapse within the first 2 years but had no risk of active TB after that. The simulation ran over a time horizon of 25 years with annual cycle length. Individuals could enter the absorbing death state at any cycle through natural or TB-related death. The two sources of death were tracked separately so that the impact of adherence on TB-related deaths could be evaluated. The simulation tracked TB-related events (incidence and deaths), healthcare costs and quality-adjusted life years (QALYs). One level of secondary transmission was modelled deterministically by applying the average number of secondary transmissions to each case of active TB, as has been done in previous studies.20 21 The model was developed using TreeAge Pro 2016 software (Williamstown, Massachusetts, USA).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Two-stage latent tuberculosis infection (LTBI) treatment outcome decision analytic model. The decision tree represents LTBI therapy outcomes based on isonicotinylhydrazide (INH) as first choice therapy followed by rifampin in cases of intolerance. The second stage was a Markov model that simulated the remaining time horizon where patients experienced a differential risk of tuberculosis (TB) reactivation based on their adherence to drug therapy.

LTBI drug therapy

In Canada, standard first-line LTBI therapy is 9 months of daily isoniazid (isonicotinylhydrazide (INH)) so it was used as the first treatment regimen in the simulation.10 Individuals were categorised into four groups of adherence to INH: completed ≥80% of doses (fully adherent), completed ≥6 months of therapy but <80% of doses, completed 3–6 months of therapy and completed <3 months of therapy. If an individual was intolerant to INH therapy, they switched to a 4-month regimen of daily rifampin (RIF). Individuals were categorised into ‘completed’ or ‘did not complete therapy’ for RIF, given the lower duration of treatment compared with INH therapy. Irrespective of the regimen used, adherence interventions were simulated to improve the proportion of simulated patients who were fully adherent to their respective drug therapy. Pooled adherence rates from large multicentre studies in low-incidence TB settings characterised the completion rates to INH (61% completion) and RIF (75% completion) therapies (table 1).

Disease inputs

Model parameters and their probability distributions, representing uncertainty about their true values, came from published literature or other secondary sources (table 1). If constructing probability distributions from the available data was not feasible, we specified plausible distributions based on expert opinion. Individual-specific adherence rates were modelled based on published studies and represented the ‘standard care’ scenario.3–5 19 Partial protective effects were given for less than full adherence for INH for 3–6 months or 6–9 months of treatment completion. There is a lack of evidence that partial RIF completion offers a protective effect, so we assumed no partial protective effect for less than full adherence.22 The lifetime risk of TB reactivation and the risk reduction of lifetime TB reactivation with INH therapy was simulated based on observations from the WHO International Union Against Tuberculosis trial.11 Risk reduction of lifetime TB reactivation with RIF was simulated based on best estimates of effectiveness from a previous study.22 We assumed that individual lifetime TB reactivation risk was constant over time.23

Economic inputs

A health system perspective was used to estimate costs. TB diagnosis and care costs were based on a Canadian national report.24 Costs of LTBI drug therapy and monitoring were provided by the British Columbia Centre for Disease Control, a centralised provincial public health agency responsible for the majority of diagnoses and treatments of active and latent TB infections in the province. Costs measured in previous years were adjusted to 2016 Canadian dollars using the Organisation for Economic Co-operation and Development consumer price indices data.25 Health state utility values for TB and LTBI states were obtained from a recent study using the SF-36 survey.26 The costs and QALYs were discounted at a rate of 3%. Willingness-to-pay (WTP) threshold of $C50 000 was used to interpret all results in the base case.27

Analytical approach

In the primary analysis, the health and economic outcomes of a ‘full adherence’ scenario were compared with the ‘current adherence’ scenario to evaluate the impact on TB cases, TB deaths and costs. As the purpose of this approach was to estimate the burden (lost opportunities) attributable to low adherence, no costs were assigned for any adherence-improving intervention. Subsequently, a hypothetical adherence intervention was simulated and compared with the current adherence scenario. In probabilistic analyses, the effectiveness of the hypothetical intervention was varied over the uncertainty distributions reported in previous studies. The relationship between a hypothetical intervention’s efficacy and maximum allowable costs of that intervention was estimated such that it remained cost-effective at the specified WTP threshold.

Evaluation of existing adherence interventions

In a secondary analysis, four existing or emerging adherence interventions identified through literature review were evaluated against standard care. A literature review was conducted to identify potential adherence interventions that could improve LTBI drug therapy adherence. Three systematic reviews summarised randomised controlled trials of medication adherence interventions for TB/LTBI or for other health conditions related or similar to TB (eg, HIV, chronic/semichronic diseases).28–30 The WHO LTBI management guidelines for low-incidence countries also reported potential interventions.31 These reviews summarised evidence largely from randomised controlled trials from multiple settings and were the most relevant sources for intervention effectiveness. Other interventions were considered but excluded because they were significantly more expensive (eg, directly observed therapy).

Four existing adherence intervention types were considered in this evaluation. The first intervention came from a review that summarised evidence of economic incentives to promote adherence to long-course LTBI or TB drug therapy.28 The incentives could include monetary rewards or other types of rewards (eg, food) for adherence. The pooled estimate for adherence incentives compared with standard care was used in the current analysis (RR=1.04, 95% CI 0.97 to 1.13). During the probabilistic simulations, the effectiveness estimates of all interventions were drawn from log-normal distributions matching the pooled point estimates and uncertainty intervals. The next adherence interventions came from a review that summarised evidence for two-way weekly text message (short message service (SMS)) adherence support.30 The SMS intervention evidence came from trials in long-term conditions such as HIV or hypertension and reported outcomes over 1 year. Based on similarities in the length of treatment observation and asymptomatic nature of the conditions in the review, we assumed SMS interventions would have a similar efficacy for the LTBI population. For the purpose of this study, the pooled effect size of two-way SMS interventions was used (RR=1.23, 95% CI 1.13 to 1.35). The third review summarised evidence for a set of adherence interventions and found enhanced adherence counselling to be effective.29 Enhanced adherence counselling involved more intensive patient education, monitoring and counselling related to the importance of drug therapy adherence. Basic adherence counselling is typically provided at prescription pick-up in British Columbia; however, this review summarised evidence of trials studying an enhancement or supplement to standard adherence counselling.29 The specific enhancements varied between the included studies but included up to three additional counselling sessions focused on the importance of adherence.29 Adherence counselling could also be enhanced through regular patient education sessions, monitoring of adherence using diaries and motivational interviewing.32–34 Once again, the pooled effect size was used to specify the input distribution used in the simulation (RR=1.09, 95% CI 1.01 to 1.15).

The WHO guidelines highlighted three studies of peer support interventions.31 In this type of intervention, an individual who has previously completed LTBI treatment would support an individual who is initiating treatment. Peers supporters could also be a friend or family member of the individual initiating drug therapy. In either case, the peer supporter receives some training on how to support the patient and provide ongoing adherence support throughout their drug therapy. The effect sizes reported in these three studies were pooled to derive the effectiveness estimate used in this analysis (RR=1.10, 95% CI 1.00 to 1.29).

Cost data were unavailable for some interventions, so empirical probability distributions could not be specified. An alternative approach was used to estimate the value of these interventions by assigning deterministic costs to each intervention and assessing the cost-effectiveness acceptability at different levels of costs. The results describe the relationship between intervention cost and the likelihood of being cost-effective (the sensitivity of each intervention to cost). The interventions were evaluated individually compared with standard care to estimate the relationship between intervention cost and cost-effectiveness acceptability (ie, probability of being cost-effective over standard care).

Sensitivity analysis

In addition to the probabilistic analysis, key assumptions were varied in the model in a series of deterministic analyses. The time horizon and WTP threshold were varied to examine their impact on the final results. The primary analysis was re-evaluated using 5-year and 10-year horizons. Based on proposed alternative WTP thresholds, the primary and secondary analyses were re-evaluated using $C20 000/QALY and $C100 000/QALY.27