SMA is a complex disease that requires a multifaceted treatment plan, including medications, physical therapy, respiratory support, and various other interventions [25, 26]. Adherence to the prescribed treatment regimen is important for improving and maintaining motor function, and can substantially impact the treatment outcomes, health-related quality of life, and economic burden of patients with SMA [16]. The recent approval of risdiplam as a daily oral medication for patients with SMA prompts the need for evidence of its associated treatment patterns in the real world. This US claims database analysis bridged that gap by assessing the real-world adherence and persistence of patients with SMA treated with risdiplam, as well as the impact of nonadherence on their health care costs.

The results indicated that more than 80% of patients with SMA receiving risdiplam were adherent to the dosing regimen, which was consistent with findings from a prior study using a national specialty pharmacy database [27]. The level of adherence to the risdiplam dosing regimen was greater compared with nusinersen, as previous studies demonstrated mean adherence rates for nusinersen ranging from 71.8 to 75.6% across SMA types 1–3 and the adherence rate dropping to 41% at 56 weeks post-treatment initiation [16, 17]. The more invasive and distressing mode of administration for nusinersen may have compromised its adherence rate.

Despite the overall high adherence to the risdiplam dosing regimen, some variation across different age groups and SMA types were observed. For example, adherence to risdiplam was observed to be more than 80% in the age groups of 0–2 years (100%), 3–5 years (90%), 13–17 years (80%), and ≥ 18 years (84.5%), but under 80% in the age group of 6–12 years (76.5%). In addition, the adherence rate varied for patients with different SMA types, with the lowest adherence for patients with SMA type 4 (75%) and the highest for SMA type 1 (100%, albeit there is only one patient in the type 1 category). Nonadherence to risdiplam may be attributed to various factors, including those related to medications (e.g., polypharmacy, adverse events, costs); potential insufficient communication between the patient/caregiver and health care provider; insurance coverage; socioeconomic factors (e.g., financial hardship); or medically needed (e.g., disease progression requiring mechanical ventilation or nutritional support) and hence discontinuing risdiplam and switching to other treatments [28]. The variation in adherence rates across age groups could be attributed to the extent of involvement of parents or adult caregivers in administering the medication, as well as potential disease progression. Caregivers of the younger age group (e.g., 0–5 years old) are often acutely aware of the potential consequences of nonadherence, such as deteriorating health conditions or hospitalization, which may contribute to greater adherence rates for this age group. Likewise, adults tend to have a heightened awareness of the consequences of nonadherence, potentially contributing to greater adherence rates in this age group (≥ 18 years). We speculate that patients between 6 and 17 years of age may have a growing self-awareness, want more autonomy, and may not always follow the daily dosing routine [29, 30]. In addition, due to progressive muscle weakness, some patients with SMA type 2 may require mechanical ventilation as they age, which could lead to discontinuation of risdiplam and a switch to other treatments, consequently with a low adherence to the treatment.

SMA is a chronic, costly disease and nonadherence may further increase the economic impact to both patients and the US health care system. Because of the need for ongoing medical management and frequent hospitalizations, patients with SMA unsurprisingly incur substantially greater health care costs than those without the disease [31]. For example, a retrospective study using data from the US Department of Defense Military Healthcare System (2003–2012) estimated that the median total expenditure for patients with SMA over 7 years was approximately $80,000 greater than those for patients without SMA, with annualized costs of approximately $48,000 versus $1800, respectively [32]. Similarly, a retrospective study using data from the Truven Health Analytics MarketScan claims database (2012–2016) found that the mean annualized total net payment for inpatient admission was $118,609 for infants with SMA compared to $58.79 for infants without the disease, and a similar trend was observed for outpatient payments ($55,538 vs. $2,047, respectively) [33].

An important finding of this study is that patients with SMA types 2, 3, or 4 who were nonadherent to their risdiplam regimen incurred greater total health care costs compared with their adherent counterparts. These results were consistent with the findings of a recent (2021) study that reported that patients with SMA who were nonadherent to nusinersen incurred greater health care costs compared with adherent patients across SMA types 1–3 [16]. Together, these findings support the substantial economic impact related to nonadherence for patients with SMA who are treated with DMTs. In addition, our study found that patients who were highly adherent to risdiplam (i.e., those with a PDC of 0.90 to 1.00) incurred $7,106 lower median health care costs than patients who were adherent to a lesser extent (PDC between 0.80 and 0.90), highlighting the relationship between medical service costs and the extent of adherence to the treatment regimen.

Multiple factors can contribute to elevated health care expenses for nonadherent patients, a primary one being that medication nonadherence can negatively impact treatment effectiveness [34]. Worsened clinical outcomes may result in increased morbidity, lower health-related quality of life, and more frequent health care resource use (e.g., increased inpatient, outpatient, ED visits) resulting in excess health care costs [35]. Nonadherent patients across SMA types 2−4 incurred higher costs for medical services compared with adherent patients. The cost differences dramatically increased with disease severity and were substantially greater for nonadherent patients with SMA type 2 (total medical service costs PPPY: $376,188), largely attributable to the need for mechanical ventilation and nutritional support. This finding underscores the critical role of treatment adherence in mitigating potentially costly complications and providing disease control, thereby reducing morbidity as well as health care resource utilization and costs.

Finally, it is worth noting that although the traditional definition of adherence as 80% PDC is generally acceptable for many chronic diseases (i.e., hypertension [36], diabetes [37]), this may not be the case for potentially rare, progressive disorders like SMA in which even minor treatment gaps may substantially impact disease management. Unfortunately, medication adherence rates for individuals with rare diseases are among the lowest for any diseases, and the rates were reported to be 58–65% in a global study of nonadherence conducted in the United States, United Kingdom, Germany, Australia, and New Zealand [38]. Few studies have examined the impact of nonadherence for individuals with rare diseases, although rare diseases have disproportionate mortality rates. For example, a population-based registry study in Italy found that although only 5% of people have been diagnosed with a rare disease, they accounted for 13% of deaths annually [39]. Poor or even conventionally ‘good’ adherence may lead to irreversible functional decline or potentially life-threatening complications for patients with SMA. More assessments are needed to better understand the optimal adherence level for patients with SMA.

The results of this study should be interpreted in the light of several limitations, some of which are common for retrospective claims database analyses. First, this study used a sample of commercially insured individuals; thus, the results may have limited generalizability to Medicaid and/or uninsured populations. Second, we defined exposure to risdiplam using the days of supply records in the pharmacy claims, which may not accurately reflect whether the medication was actually taken by the patient. Third, as administrative claims data are collected for payment rather than research purposes, there may be billing inaccuracies and missing data related to miscoding of medical diagnoses. Thus, the analysis is vulnerable to coding inaccuracies. Fourth, there is a potential risk of misclassification of SMA types based on an established algorithm as such information is not available in health insurance claims data. In addition, SMA is a rare disease, which limited the sample size of eligible patients for this study. Only one patient with SMA type 1 receiving risdiplam was captured. One possible reason is that risdiplam was first approved for adults and children 2 months and older on August 7, 2020, and later for children under 2 months of age (May 31, 2022), which was roughly 1 month before the end of the available data period, limiting this patient population for study inclusion. In conclusion, the analyses conducted in this study were descriptive in nature and therefore causal relationships could not be ascertained. Future studies with a larger sample size and longer follow-up period may allow for a statistical comparison of health care costs between adherent and nonadherent patients while adjusting for age and other confounding factors.