Cutting Off Benzodiazepines Doesn't Lower Hip Fracture Risk

BOSTON, Jan. 16 -- Limiting access to benzodiazepines doesn't appear to reduce the risk of hip fractures for older patients, researchers here reported.

After New York enacted new rules that had the effect of limiting prescriptions for benzodiazepines to older Medicaid patients, their use dropped by nearly two-thirds, but there was no change in hip fracture rates, reported investigators in the Jan. 16 issue of the Annals of Internal Medicine.

There were similarly no changes in hip fracture rates in neighboring New Jersey, where there were no changes in rules governing benzodiazepine prescription or in their use, reported Anita Wagner, Pharm.D., M.PH, Dr. P.H., of Harvard Medical School and colleagues.

"It is very challenging to answer the question whether benzodiazepines cause hip fractures, Dr. Wagner said. "People who get benzodiazepines, such as chronically ill elderly patients with dementia, have conditions, like dementia, that can cause hip fractures -- and their hip fractures may not be due to their benzodiazepines."

Benzodiazepines can cause confusion and balance problems in older people, leading to increase risks for falls and fractures. About one-third of states restrict access to benzodiazepines by various means, and the drugs are excluded from coverage under the Medicare Part D drug benefit, the authors noted.

When New York State started in 1989 to require physicians to use special serially numbered, triplicate forms to prescribe benzodiazepines (to allow for closer monitoring of the drugs), prescriptions for the agents dropped by 55% among Medicaid recipients and by 44% among all residents of the state.

In contrast, the demographically similar state of New Jersey did not make any changes to its policies on benzodiazepine prescriptions, and use of the drugs there did not change, making it ideal for comparison purposes.

"Our hypothesis, like that of policy makers and other researchers," the authors wrote, "was that the sudden, large, sustained decrease in benzodiazepine prescribing in New York would result in a decrease in the incidence of hip fractures, particularly among those at highest risk for hip fractures, women who use benzodiazepines, while hip fracture rates would not decrease in New Jersey."

The investigators conducted a "quasi-experiment" in which they examined changes in outcomes among Medicaid recipients in New York versus their peers in New Jersey following the start of New York's triplicate-prescription form. Pharmacists are required to send one copy of the form to state health authorities, allowing the authorities to monitor drug prescription, dispensing, and use.

The authors looked for changes in hip fracture rates in a cohort of 51,529 Medicaid patients in New York, and 42,029 in New Jersey, all older than 65. The outcomes measurements were benzodiazepine prescribing and hip fracture ratios adjusted for age and eligibility category.

They found that the triplicate prescription policy caused benzodiazepine use in women on Medicaid to drop by 60.3% (95% confidence interval -66.3%, -54.2), and in men to drop by 58.5% (-64.3%, -52.8%). At the same time, use of the drugs in New Jersey remained stable.

They also found, however, that the precipitous fall in the use of the medications had no apparent effects on hip fracture. Among women in both states who had received benzodiazepine prescriptions before the new policy was put in place, the hazard ratio for hip fracture was 1.2 (95% CI, 0.80-1.66) before the policy change, and 1.1 (95% CI 0.80-1.40) after (P=0.70). Similarly, among women who didn't take the drugs, the pre-policy hazard ratio was 1.3 (95% CI 1.07-1.68) compared with an after-policy hazard ratio of 1.1 (95% CI 0.90-1.23, P=0.08).

Among men who had received benzodiazepines before the new law went into place, the risk for fractures pre-policy was 0.8 (95% CI, 0.33-1.85) and the risk after the policy went into place was 1.1 (0.55-2.09, P=0.56). Among non-users, the hazard ratio before the 1989 change was 1.1 (95% CI 0.67-1.69), and after the change was 1.3 (0.94-1.86, P=0.46).

The authors suggested that the most likely explanation for their findings, which diverge from previous studies suggesting a link between benzodiazepines and hip-fracture risk, is that the previous studies did not adequately control for confounding factors that are not included in claims data, including body mass index, smoking, activity and daily living score, cognitive impairment, and a physical impairment scale score.

"The challenge of disentangling the effects of benzodiazepines from other causes of hip fractures in the elderly is especially concerning when study results are used to guide policies that restrict access to medicines for huge populations," said senior author Stephen Soumerai, Sc.D., a professor of ambulatory care and prevention at Harvard Medical School.