Model overview

A model for decision analysis was created to assess the cost-effectiveness of olanzapine ODT versus SOT. The structure of the model used in the decision analysis represented the pathway of a patient with schizophrenia receiving either the ODT and SOT formulations of olanzapine, risperidone, and aripiprazole. Patients were classified as compliant, partially compliant, or non-compliant with the treatment, and their adherence level was taken as the determinant as to whether patients may either remain stable or have a relapse, which may or may not necessitate hospitalization.

The model used in the decision analysis included six treatment groups that involved three commonly used atypical antipsychotics, including olanzapine, risperidone, and aripiprazole, that were included in both ODT and SOT formulations in the model. Additionally, the model incorporated three treatment groups that included ODT and SOT formulations of different antipsychotics as a group [risperidone (ODT + SOT), aripiprazole (ODT + SOT), and olanzapine (ODT + SOT)]. The model was designed to simulate the standard of care process for schizophrenia patients for a period of one year, taking into account the dynamic nature of their condition. The model considered several input parameters, such as adherence rates, relapse with or without hospitalization, health state utilities, adverse events, resource utilization in the healthcare system, and direct healthcare costs. The simulation involved 1,000,000 patients and led to the prediction of significant clinical outcomes such as quality-adjusted life years. As the model was developed for Morocco, all Costs are denoted in Moroccan Dirhams (MAD) using 2022 values. Additionally, based on the cost data received from Morocco, ODT formulations of antipsychotic medications were more expensive than their SOT counterparts. The approach used in the model was intent-to-treat, where all direct medical costs estimated were attributed to the patient’s initial treatment received. Additionally, adverse reactions that may be reported by patients during treatment, such as extrapyramidal symptoms (EPS), significant weight gain, or diabetes we also accounted for, in the model. Figure 1 presents an overview of the model.

Fig. 1

Illustration of the model’s framework

Key clinical and economic input values

Given a lack of available clinical data estimates for Morocco, we had to use data published in a cost effectiveness study by Ascher-Svanum et al. [22] to obtain values for key clinical endpoints. In cases where data could not be obtained from peer-reviewed articles, expert opinion of psychiatrists in Morocco was utilized to understand treatment patterns and resource utilization. The model assumed that ODT formulations of the three antipsychotics assessed were equal in efficacy and safety to their respective SOT formulations based on all clinical input parameters, with the exception of greater adherence in patients receiving ODT which was in line with published comparative data.

Adherence levels

As there is no publicly available or published data on adherence rates, we used the following data from published literature as input parameters in the model. Consistent with prior research, adherence levels were categorized based on the medication possession ratio (MPR) as: adherent (MPR ≥ 80%), partially adherent (MPR ≥ 60%, ≤ 80%), and non-adherent (MPR ≤ 60%). The adherence rates used in the model are shown in Table 1 along with the data source.

Table 1 Adherence rates by medicationRelapse rates

Table 2 outlines the assumptions used in the study for the probability of the first relapse that requires hospitalization and relapse not requiring hospitalizations by adherence category for each medication.

Table 2 Relapse rates requiring and not requiring hospitalizationTreatment-emergent adverse events

To simulate the effects of antipsychotic treatment, the model requires assumptions about the likelihood of patients experiencing various types of treatment-related adverse events, such as extrapyramidal symptoms (EPS), clinically significant weight gain (defined as an increase in weight of at least 7% from baseline weight), and diabetes. Table 3 outlines the initial assumptions about treatment-related adverse events for each medication and their data sources.

Table 3 Treatment emergent adverse eventUtility and quality-adjusted life years

The model’s starting utility values for the nine potential scenarios involving levels of adherence and relapse status are presented in Table 4.

Table 4 Utility values for health statesMedication cost

Daily dose amounts and medication expense are frequently correlated. We used the daily dose levels reported by psychiatrists in Morocco to ensure that schizophrenia patients receive equivalent medication dosages. Table 5 shows the 2022 net wholesale price, which reflects the baseline model assumptions for dosing and expense for each drug. It demonstrates that ODT antipsychotics are more expensive in Morocco as compared to SOT equivalents.

Table 5 Economic input parameters; medication costsHealthcare resource utilization

Table 6 provides information on the assumptions made regarding the utilization of eight different types of medical services for 5 different patient results, along with the sources of data used to generate this information.

Table 6 Healthcare resource utilizationHealthcare resource cost

Table 7 presents the initial expenses of each healthcare resource used. The costs of each unit were adjusted for inflation to reflect the value of the Moroccan MAD in 2022, employing the section of the consumer price index that pertains to healthcare services.

Table 7 Healthcare resource costModel outcome measuresClinical outcomes

The model estimates three critical clinical outcomes: the percentage of patients who experience an outpatient relapse, those who experience an inpatient relapse, and those who do not experience either an outpatient relapse or an inpatient relapse (i.e., stable).

Economic outcomes

The model reports mean total direct healthcare costs for ODT formulation along with ODT plus SOT formulation as a group for the all three antipsychotic drugs.

Cost-effectiveness information

The cost per one QALY gained for each medication is the main measure of cost-effectiveness. Additionally, the model computes incremental cost-effectiveness ratios (ICERs), which are calculated by dividing the cost variation by the variation in the proper measure of effectiveness.

Sensitivity analysis

A one-way sensitivity analysis (OWSA) was conducted by using sequential bifurcation, a process that iteratively samples inputs and assesses the impact of each input against a pre-determined cost threshold value, to determine what variables affecting total treatment costs warrant focus during sensitivity analyses. Additionally, to test the robustness of the model concerning uncertainty in model input parameters, a probabilistic sensitivity analysis (PSA) is performed using a second-order Monte Carlo simulation with 1000 iterations. Each key model parameter is given a theoretical probability distribution in this analysis. A random number generator is used to draw parameter values from each distribution, and these values are run through the model to generate a cost-effectiveness scatter plot.