Adherence with tobramycin inhaled solution and health care utilization

Over the past 20 years, median life expectancy for patients with cystic fibrosis (CF) has increased; recent data from the Cystic Fibrosis Foundation Patient Registry demonstrated a median survival of 37.4 years among patients in the United States (US)[1]. These improvements in survival are associated with earlier diagnosis, more active treatment of pulmonary and digestive complications, and a growing number of effective therapies [2]. One of the consequences of these medical advances is the increased duration and complexity of daily treatments for CF [3]. As research in other chronic illnesses has shown, both time and complexity of daily regimen are associated with lower rates of adherence [4], and these observations may also be true for the CF population[5].

In CF, bacterial inflammation and infections of the lungs are common, particularly those caused by the pathogen Pseudomonas aeruginosa (P. aeruginosa). By the age of 18, approximately 80% of individuals with CF test positive for P. aeruginosa[1, 6]. Individuals with CF and P. aeruginosa infections have twice the risk of hospitalization for acute respiratory exacerbations and twice the risk of death as those without P.aeruginosa infections[7, 8]. The aminogylycosdies tobramycin, gentaicin and colistin are active against P. aeruginosa and these drugs are administered by inhalation, although only tobramycin inhalation solution (TIS) is approved for inhalation. The Cystic Fibrosis Pulmonary Guidelines: Chronic Medications for Maintenance of Lung Health recommend the chronic use of TIS if P. aeruginosa is persistently present in airway cultures, regardless of the severity of the disease, although the recommendation is strongest in the case of moderate to severe disease [9].

Reported rates of prescribing TIS to eligible patients are highly variable, ranging from 47-90% [3, 10, 11]. Estimates from the CF Foundation patient registry suggest that nearly 58% of individuals with CF and P. aeruginosa have had TIS prescribed at least once, although there are no estimates of chronic use[9]. Other estimates of TIS use have been limited by the lack of data beyond initial receipt of the medication and the use of self-reported questionnaires, which can influence adherence estimates[3, 4, 12, 13].

In the U.S., the majority of patients with CF receive their health care through a network of over 120 CF care centers and affiliated programs that are located nationwide. CF centers provide multi-disciplinary care teams of physicians, nurses, social workers, and respiratory therapists. Approximately 60% of CF patients have their health care services and products paid for by employer-based health insurance that is offered to employees and their dependents as part of the employment benefit package[1]. Detailed insurance claims are generated for every prescription drug, hospitalization, and physician service received by patients with CF and thus claim-based databases are one of the most comprehensive sources for assessing the overall health care utilization of this patient population in the U.S. [14].

The objectives of this study were to: 1) measure adherence with TIS as documented in prescription claims data and 2) assess the relationship between use of TIS and health care utilization, particularly hospitalizations.

To our knowledge, this is one of the first studies to assess the link between TIS use in CF patients and health care utilization as captured in health claims data. In this study of 804 individuals with CF, 56% were identified as having P. aeruginosa, although only 6% of those with P. aeruginosa were dispensed at least 4 cycles during the year of observation. These findings are similar to those found in a previous study of 357 individuals with CF and P. aeruginosa, where only 4% were utilizing chronic intermittent TIS according to guidelines [21], although there was overlap in the study samples.

Rates of aerosolized antibiotic use in CF patients have been previously reported. In the CFF Patient Registry, the national average rate of reported use (based on provider prescribing) was 60.8% (range 17.2-94.7%) in 2001 and 61.8% (range 0-100%) in 2006 [22]. However, these data were not analyzed by P. aeruginosa status. Analysis of the Epidemiologic Study of CF database showed that 73.1% of patients with P. aeruginosa and 33.8% of patients without P. aeruginosa used inhaled tobramycin from 2003-2005 and that 49.7% vs. 15.9% of all patients received inhaled tobramycin for chronic vs. exacerbation use, respectively [23]. Neither of these previous analyses assessed dispensing or refill history or if inhaled tobramycin was used according to practice guidelines, and both assessed all formulations of aerosolized tobramycin.

In the pivotal clinical trial of TIS, Ramsey et al. noted that subjects in the tobramycin group were 26% less likely to be hospitalized and 36% less likely to be treated with intravenous antibiotics than subjects in the placebo group [24]. In the current study, CF subjects with high utilization (≥ 4 cycles per year) had fewer hospitalizations than those subjects with medium and low utilization; the risk of hospitalization among those with high utilization was decreased by 60% compared with those taking 2 or fewer cycles per year. Thus, our findings are consistent with the clinical trial results, and suggest that TIS use in routine clinical care and not within a clinical trial may also be associated with a reduction in hospitalizations. This finding deserves further study as the sample size of CF patients with high utilization was modest (n = 54) and many factors besides TIS adherence contribute to hospitalization risk.

We noted that while median outpatient costs, excluding drug costs, were lower for high users, median outpatient drug costs were higher. Previous studies have sought to examine whether the cost of TIS can be partially offset by reductions in the use of other health care resources. LeLorier et al. and Iles et al. have shown that the use of TIS could potentially reduce the use of health care services, particularly hospitalization and intravenous antibiotic use (29, 30). Furthermore, while not measured by studies to date, decreased hospitalization is likely to result in additional cost saving resulting from fewer disruptions of work and school as well as improved quality of life for patients and caregivers

Poor adherence to TIS, as measured in this study may be due to several factors. Patients may not have a complete understanding of TIS and its risks and benefits and they also have several other medications and treatments to complete each day. In addition to the treatment burden, use of TIS may be limited because of barriers such as taste, portability of the treatment or inability to pay the coinsurance (31). Although TIS is recommended for patients with P. aeruginosa infection, prescribing patterns are not consistent nationally, and TIS is likely increasingly being used both as part of early eradication protocols and to treat CF pulmonary exacerbations. Thus, low utilization or fewer annual cycles may reflect these practice patterns.

There are some limitations to this study. This study used pharmacy claims records to measure adherence, and thus our analysis assessed rates of medication acquisition rather than medication exposure. However, research has demonstrated predictive validity for measuring the cumulative exposure of medications with acquisition data[25]. It should be noted, though, that studies using electronic data capturing methods have found a range of adherence rates with some reporting higher levels of TIS adherence in selected populations[26]. Second, P. aeruginosa infection and the rate of failure to thrive may be under-coded due to lack of financial incentive to add these diagnoses to billing records. Additionally, the DCG/HCC model has not, to our knowledge, been previously applied to a CF population. Thus, its performance as validated in general populations may vary from what was observed in our study population. However, there is value to applying a standard risk-adjustment methodology to minimize the differences in comorbidity levels observed in the sample. In this dataset we do not have an objective measure of disease severity, such as forced expiratory volume in one second (FEV₁), to correlate lung health with TIS use. Further, it should be noted that our measure of adherence does not assess all aspects of physician prescribing patterns (e.g. use of generic tobramycin). Also, this study population may not be generalizable to the American CF population; although the proportion of males and adult patients is comparable to that reported to the Patient Registry in 2007, the distribution of geographic regions is not comparable to Registry data (1). Fewer patients from the Northeastern United States were identified for inclusion in the current study, although the impact of geographic region on prescribing patterns for TIS and access to care are unknown. This dataset reflects patients with private health care insurance and findings may not be generalizable to patients on public assistance. Finally, this analysis was limited to one year of observation and did not assess utilization patterns after that period.

The value of a dataset derived from administrative claims data may be unfamiliar to CF clinicians, however, such data are appropriate for assessing many facets of quality of care. One particular strength of these data is the ability to capture medication adherence as measured through prescription dispensing; in this study, we found that persistent use of TIS was quite low and deserves further investigation. We are currently conducting a randomized controlled trial across 19 CF Care Centers (iCARE: I Change Adherence and Raise Expectations Study, NCT01232478) to identify effective strategies that improve adherence in adolescents with CF.

High utilization of TIS was infrequent yet associated with a decreased risk of hospitalization. However, the low utilization of TIS in subjects with P. aeruginosa noted in this study is of concern. These data suggest that CF patients may not be using TIS as recommended by the CF Pulmonary Guidelines. Future studies to address ways to improve adherence are warranted to ensure patients can benefit from their chronic treatments.