Formulary Exclusion of Esomeprazole: Impact on Healthcare Costs and Utilization | Page 3
Published Online: August 15, 2013
Berhanu Alemayehu, DrPH; Joseph A. Crawley, MS; Xiongkan Ke, MS; and Marta Illueca, MD
In patients taking a PPI for risk reduction of NSAID-associated gastric ulcer, average adjusted total and GIrelated costs incurred by patients who switched to another PPI were numerically higher compared with those who remained on esomeprazole, but the differences were not statistically significant ($25,233 vs $24,348 and $3131 vs $2885, respectively; Figure 2D). Patients who switched to another PPI had numerically higher average total prescription costs compared with those who remained on esomeprazole ($6010 vs $5799). Compared with patients who remained on esomeprazole, patients who switched to another PPI incurred significantly higher (P = .0425) average PPI prescription costs ($1234 vs $1154).
Mixed Cohort Group
From the pre-exclusion to the postexclusion period, average total and average GI-specific healthcare utilization increased in the mixed PPI group and decreased in the esomeprazole group. For the mixed versus the esomeprazole group, the relative change in average total healthcare utilization was 19.2% for inpatient visits, 5.7% for outpatient and emergency visits, 5.1% for office visits, and 6.2% for other visits. Similarly, for the mixed versus the esomeprazole cohort, the relative change in average GI-specific healthcare utilization was 26.3% for inpatientvisits, 21.7% for outpatient and emergency visits, 26.3% for office visits, and 14.4% for other visits. From the preexclusion to the postexclusion period, the percent change in average number of prescriptions was also higher in the mixed group compared with the esomeprazole group, with a relative change of 7.2% in total prescriptions and 16.7% in PPI prescriptions.
In the total study population, patients in the mixed group incurred significantly higher (P <.0002) average adjusted total costs, total prescription costs, PPI prescription costs, and medical costs per patient per year compared with patients who remained on esomeprazole therapy in the postexclusion period ($22,793 vs $21,131, $4494 vs $4075, $1234 vs $1079, and $3503 vs $2965, respectively). Compared with patients who switched to another PPI, patients in the mixed cohort group incurred significantly higher (P <.0145) average adjusted total prescription costs and PPI prescription costs per patient per year ($4494 vs $4398 and $1234 vs $1194, respectively). However, patients in the mixed cohort group and patients who switched to another PPI incurred similar total costs and medical costs ($22,793 vs $22,237 and $3503 vs $3379, respectively).
In a recent review of the role of comparative effectiveness research in helping to reduce healthcare costs and improve the quality of care, Brixner and Watkins noted, “When assessing the impact of cost, it is important to look beyond drug costs to the overall impact on treatment costs, including cost offsets that may occur through improved health or decreased morbidity.”20 This study examined patterns of healthcare utilization and cost implications for 12 months following the exclusion of esomeprazole from the United Healthcare formulary in patients who had been taking esomeprazole for EE, GERD, or risk reduction of NSAID-associated gastric ulcer. Findings showed that patients who switched to another PPI incurred higher healthcare utilization and significantly higher average adjusted total medical, GIrelated medical, total prescription, and PPI prescription costs per patient per year compared with patients who continued on esomeprazole. These findings suggest tha tformulary exclusions can lead to unexpected higher overall healthcare utilization and costs.
Findings from previous pharmacologic studies suggest that esomeprazole more effectively controls acid and promotes11-14 and maintains15-17 healing of certain conditions (eg, EE) than other PPIs. These findings support the present findings and provide a potential explanation for the lower healthcare utilization and costs (ie, office visits, ED visits, prescription costs) associated with using esomeprazole for EE or GERD compared with other PPIs. In addition, fewer PPI prescriptions were used in the group that remained on esomeprazole than in the group that switched to another PPI, providing another potential explanation for the lower costs associated with continued use of esomeprazole compared with switching to other PPIs.
It also is possible that this formulary exclusion led to negative health outcomes by affecting not only GI diseases, but also other comorbidities. Findings from the present study showing that patients who switched to another PPI experienced greater increases in overall medical use and expenditures than in GI-specific medical use and expenditures support this hypothesis. Gastroesophageal reflux disease is associated with several non-GI comorbidities. 21 The most common comorbidities associated with GERD include hypertension, hypercholesterolemia, obesity, diabetes, and depression.21 Patients taking a PPI to reduce the risk of NSAID-associated gastric ulcer could be taking the NSAID for various diseases, which mayalso be associated with comorbidities (eg, osteoarthritis including hypertension and coronary artery disease).22 As noted in the previous study assessing healthcare utilization 6 months after esomeprazole formulary exclusion, further research into the effect the formulary exclusion may have on healthcare costs and utilization associated with non-GI comorbidities is warranted.18
Analyses of the subgroups of patients with EE, GERD, and patients taking a PPI for risk reduction of NSAIDassociated gastric ulcer also showed lower adjusted total and GI-related healthcare costs per patient per year in patients who remained on esomeprazole compared with those who switched to another PPI. Differences between the patients who remained on esomeprazole and those who switched to another PPI were significant in patients with EE or GERD, but not significant in the patients taking a PPI for risk reduction of NSAID-associated gastric ulcer. This may be explained by the small sample size ofthat particular group (n = 757 patients). Further evaluation with a larger pool of patients receiving a PPI for risk reduction of NSAID-associated gastric ulcer is needed to confirm these observations.
PDF is available on the last page.