BackgroundPeople with diabetes who inject insulin with pen devices may reuse the pen needles (PNs), a practice that can cause PN tip deformity, breakage, and contamination, and that is associated with lipohypertrophy and injection-related pain. ObjectiveThis retrospective study aimed to estimate the extent of PN reuse among people with diabetes in 2 insured populations in the United States. MethodsUsing claims data for Commercial Fully Insured (CFI) and Medicare Advantage (MA) populations from 1-Oct-2018 to 31-Dec-2022, we identified adults with type 1 or type 2 diabetes (T1D/T2D) who had ≥ 1 claim for PNs and ≥ 2 claims for insulin from 1-Jan-2019 to 31-Dec-2021, with continuous medical/pharmacy eligibility for 3 months before first claim and 1 year after (follow-up). Those receiving hospice or palliative care or using mail order prescriptions were excluded. We compared actual annual fill rate of PNs with expected fill rate (assuming single use) according to prescribed insulin regimen. Whether the annual actual-to-expected ratio for PN numbers equaled 1 was evaluated using sign tests with 2-sided P values. ResultsMedian annual actual-to-expected ratios ranged from 0.41 (T1D basal+prandial cohort) to 0.82 (T2D basal cohort; all P < 0.001) in the CFI population (N = 10,854), and from 0.55 (TID basal + prandial) to 1.10 (T2D basal and basal + prandial; P = 0.382-< 0.001) in the MA population (N = 32,495); medians were 0.34 and 0.55 for 4 expected T2D basal + prandial injections/day in CFI and MA populations, respectively (P < 0.001). Annual actual-to-expected ratios were < 1 for 62% and 47% of CFI and MA populations, respectively. An estimated 2%-27% and 0%-17%, respectively, depending on insulin regimen, had inadequate supplies of PNs suggesting that PNs could have been used ≥ 5 times. ConclusionThese findings highlight the need for educating people with diabetes about reasons for avoiding PN reuse and the key role that pharmacists can play in providing this information and adequate supplies of PNs.