MEDICAID LIMIT ON NUMBER OF PRESCRIPTIONS PER MONTH RAISES RISK OF MORE NURSING HOME ADMISSIONS, NEW HAMPSHIRE STUDY BY SOUMERAI AND AVORN FINDS

New Hampshire Medicaid's policy of limiting the number of reimburseable outpatient prescriptions per month had a "clearly observable effect" of increasing nursing home admissions of "frail, low-income, elderly patients," according to a study by Stephen Soumerai, ScD, and Harvard Medical School colleagues including Jerry Avorn, MD. Reported in the Oct. 10 issue of The New England Journal of Medicine, the study examines 36 months of Medicaid data to determine the effects of New Hampshire's experiment with a three- drug payment limit for Medicaid patients between September 1981 and July 1982. New Jersey, which had no drug-reimbursement limits in the 36-month period (July 1980 to June 1983), was used as a control. The researchers examined data on 411 New Hampshire patients and 1,375 New Jersey patients who, before the cap was imposed, had averaged at least three prescriptions per month including at least one maintenance therapy for a chronic illness (diabetes, heart disease, pulmonary disease and asthma seizures, or conditions requiring anti-coagulants). All patients were over age 60. In the year prior to the imposition of state limits on outpatient drug prescriptions, the proportions of patients entering nursing homes was similar in each state: 2.3% in New Hampshire and 2.1% in New Jersey. By the end of the 11-month period, however, 10.6% of the New Hampshire patients were in nursing homes, compared to 6.6% in New Jersey, the article says. When New Hampshire removed the prescription cap and switched to a $ 1 copayment system, the increased risk of nursing home admission subsided, although those already in the nursing homes remained there. During the cap period, drug utilization fell by 35%. Patients who were taking medications for three or more conditions before the cap went into effect were at an even greater risk in New Hampshire, the study found, with 14.4% in nursing homes compared to 6.2% in New Jersey at the time the cap was lifted. "The cap's adverse effect was most pronounced for the patients who were most disabled," the researchers said. Hospitalization rates for patients using medications for more than three conditions at baseline showed an increasing trend in New Hampshire during the cap period but did not reach statistical significance. "This trend disappeared when the cap was replaced with the $ 1 copayment policy," the article says. The researchers noted that the "increases in the rate of hospitalization may have been too low to be measured against a high background rate in a chronically ill population." Two factors could explain the increase in nursing home admission rates, the researchers noted: "declining health" or "the desire to maintain the use of essential medications because the three-drug limit did not apply in long-term care facilities." The "increase in nursing home admissions among the patients at highest risk suggests that the loss of medications could have exacerbated preexisting medical problems," the article says. However, "given case reports by New Hampshire Legal Assistance of several patients who were transferred to nursing homes to avoid the policy's effect, the cap probably precipitated nursing home admissions for financial reasons as well." The article notes that the estimated 174 person-month increase in nursing home time seen in New Hampshire during the study period cost the state $ 310,745 at its daily Medicaid reimbursement rate of $ 59. "This underestimates the true cost," the article adds, since it assumes no other expenses, such as physician costs, and no time in nursing homes after the study period. All told, the "unintended costs" of the program may have been "well over the estimated statewide savings of $ 300,000 to $ 400,000 achieved by the cap." "At present, about one fourth of state Medicaid programs have limits on drug reimbursement in effect. Our findings raise questions about the clinical and economic wisdom of such policies," the researchers concluded. "The challenge for researchers and policy makers is to discover which cost- containment methods are most efficient in reducing ineffective care while preserving access to forms of medical technology that benefit both individual patients and society as a whole." Addressing an Oct. 9 seminar at FDA, Avorn commented that the three-drugs-per-month limit described in the study "is an attempt to change prescribing practices that is not rooted in any kind of pharmacology or clinical insight, but rather it's the pointy- headed green-eye shade folks telling us what to do -- which I think results in less good care and probably does not even save money." During a review of various strategies for improving prescribing practices, Avorn listed the New Hampshire program among "things that do not work." Avorn is widely known as a pioneer of "counterdetailing" programs designed to counterbalance information provided by ads and company detail reps. Somewhat ironically, Avorn noted that the NEJM article might be seen as the "ultimate" pro-industry study on the cost effectiveness of drug therapy. Patients in the study "get institutionalized in nursing homes and they don't come out again, no big surprise," Avorn remarked. "But when you think about the economics of this -- and again I guess the industry now thinks I'm a good guy and not a bad guy -- it is the ultimate study that the industry has been hoping someone would do showing that drugs are really cost effective. Namely, you give people a little bit of Lasix and [Digoxin] and lo and behold, they don't have to be institutionalized." Avorn told the FDA seminar that he is hoping to publish another journal article within the next few months describing effects of counterdetailing on various psychoactive drugs used in a nursing home. The study compared six control nursing homes to six experimental facilities. In the latter, researchers conducted face-to-face discussions with prescribers about appropriate drug use -- a critical component of counterdetailing. Due to nurses' key role in such facilities, the researchers also counterdetailed to nurses and aides. Avorn also stressed the importance of providing "behavioral alternatives" to the less-desired drug utilization, which included explaining sleep hygiene, suggesting clinical workups for certain patients and providing a list of preferred drugs to use as needed. "In essentially every case we were able to show a reduction in utilization of the drugs in the desired direction," Avorn reported. One of the drugs the researchers "went after" and which decreased in use was diphenhydramine which is a "very common and popular [sleeping pill] in nursing homes. Geriatric conventional wisdom holds that it's not a good sleeper because its so anticholenergic." The study also found that the "nonrecommended" benzodiazepines, "particularly the long half-life ones -- Lithium, Valium, Dalmane -- went down [in use]." Avorn said. "And the recommended ones even went down, in part because the message was not use these and not those, but rather, think hard about whether this person needs a sedative medication [at all]." Regarding clinical outcomes, Avorn said that "very careful assessment of behavior as well as the context of what's going on in the nursing home revealed that there was no decrement in that regard." For example, in giving patients a memory assessment test originally developed for Alzheimer's patients, the researchers found that memory performance remained stable or improved for 79% of patients in the experimental homes versus 49% in the control homes. In addition to directing Harvard's Program for the Analysis of Clinical Strategies, Avorn is working with Medco Cost Containment Services to develop counterdetailing programs, particularly one to be used by Massachusetts Blue Cross and Blue Shield ("The Pink Sheet" May 20, p. 19).