Refill Patterns of Atypical and Conventional Antipsychotic

Document Sample
Refill Patterns of Atypical and Conventional Antipsychotic Powered By Docstoc
					Refill Patterns of Atypical and
Conventional Antipsychotic Medications
at a National Retail Pharmacy Chain
Mark Vanelli, M.D., M.H.S.
Philip Burstein, Ph.D.
Joyce Cramer

This study examined whether pa-
tients who received prescriptions
for conventional and atypical an-
                                              F    or patients with chronic mental
                                                   illness, the introduction of atypi-
                                              cal antipsychotics has opened new
                                                                                          agents. Patients who received a new
                                                                                          prescription for a noninjectable form
                                                                                          of an agent were included in the study,
tipsychotics in routine outpatient            possibilities for reintegration through     resulting in an analysis sample of
care continued medication ther-               meaningful work and community liv-          25,301 patients. The generic and
apy. Prescription refill records of           ing. However, the likelihood of pa-         brand formulations of each agent were
more than 25,000 patients from a              tients’ sustaining such improvements        analyzed. Because some patients re-
national retail pharmacy chain                may depend on their adherence to            ceived a prescription for more than
were surveyed during a full                   prescribed medications for months or        one antipsychotic, a total of 26,447
eight-month period. At the start              years in an unsupervised community          courses of therapy were analyzed.
of the ninth month, 48 percent of             setting.                                       Of the 17,067 initial prescriptions
the patients taking conventional                 Treatment guidelines suggest a           for atypical agents in our database,
agents were continuing therapy.               year of antipsychotic therapy for pa-       290 were for clozapine, 8,785 for
The overall rate for those taking             tients who are experiencing their first     risperidone, 1,662 for quetiapine, and
atypical agents was 44 percent;               psychotic episode and at least five         6,330 for olanzepine. Of the 9,380 ini-
however, the rate for those tak-              years of maintenance therapy—if not         tial prescriptions for conventional
ing clozapine was 71 percent.                 indefinite therapy—with antipsy-            agents, 4,481 were for haloperidol,
Atypical agents, despite their im-            chotics for patients who have experi-       1,211 for trifluoperazine, 1,623 for
proved side effect profile, were              enced multiple episodes (1). Yet the        thiothixene, 829 for fluphenazine,
not associated with higher rates              psychiatric literature reveals little       and 1,236 for perphenazine. Overall,
of continued treatment. A highly              about the duration and variation of         fewer than 5 percent of the patients
supervised medication adminis-                adherence to antipsychotic medica-          switched to another antipsychotic
tration process, frequent patient-            tion under conditions of routine out-       during the study period.
provider contact, and favorable               patient care.                                  The endpoint of the analysis was
patient selection may help explain               This study examined the medica-          medication persistence, that is, the
the higher rate among patients                tion adherence of patients in the           proportion of patients who continued
taking clozapine. (Psychiatric Ser-           community who obtained their initial        with therapy. Medication persistence
vices 52:1248–1250, 2001)                     prescription for an antipsychotic           was defined as a patient’s having a sup-
                                              medication at a national retail phar-       ply of medication in hand at a particu-
                                              macy chain. Our first goal was to doc-      lar point in time. Treatment persist-
                                              ument the pattern of nonadherence           ence was evaluated for each patient at
Dr. Vanelli is chief medical officer at Ad-   over an eight-month period. Our sec-        30-day intervals from the patient’s
heris Incorporated, 400 West Cummings         ond goal was to compare differences         starting date for an eight-month peri-
Park, Suite 3050, Woburn, Massachusetts       in adherence for atypical and conven-       od or until the end of the study period.
01801 (e-mail,         tional agents and among individual          Because patients who entered the
He is also a consultant with BraunCon-
                                              agents.                                     study in later months had less time
sulting in Boston. Dr. Burstein was for-
                                                                                          during which they were at risk of dis-
merly with Adheris and is currently with
Blue Cross–Blue Shield of Massachusetts       Methods                                     continuing therapy, the denominators
in Boston. Ms. Cramer is with the de-         The database for this study consisted       used for evaluating persistence de-
partment of psychiatry at Yale University     of all refill records from a national       creased with each successive month.
School of Medicine in New Haven, Con-         pharmacy chain from September 1998             By this method, patients were con-
necticut, and is a consultant to Adheris.     through May 1999 for the selected           sidered to be continuing medication
1248                                                                          PSYCHIATRIC SERVICES   o September 2001 Vol. 52 No. 9
therapy at the end of a given month if       Figure 1
they had sufficient medication from a        Percentage of patients who adhered to medication therapy, by type of antipsy-
previously purchased refill to carry         chotic agent
them past the end of that month or if
they subsequently purchased another                                           100
refill. For example, if a patient pur-
chased an initial prescription and re-                                         90

                                              Percentage continuing therapy
turned for the next refill during
month 8 of therapy, he or she was still
classified as continuing therapy for
the entire eight-month period. Thus
patients classified as continuing ther-                                        70
apy after eight months could have ob-
tained as little as two months’ worth                                          60
of medication. Persistence, as defined                                                         Clozapine
here, simply reflects a patient’s ten-                                         50              Conventionals
dency to continue with therapy; it                                                             Atypicals
does not reflect whether a patient ac-
tually took the medication, nor does it
                                                                                                   2     3     4     5       6      7       8       9
measure the clinical benefit received.                                                 Initial
   Chi square analyses were used for                                                prescription                   Start of month
comparisons between groups.

As shown in Figure 1, persistence            were taking risperidone (519 of                                        norm (3). Clozapine was unique
with therapy was greater among pa-           1,473) (χ2=41.37, df=1, p<.001), 40                                    among the agents we studied in that
tients who were taking conventional          percent of those who were taking                                       patients who were treated with this
agents than among patients who were          olanzepine (413 of 1,026) (χ2=39.76,                                   antipsychotic were required to partic-
taking atypical agents. At the start of      df=1, p<.001), and 33 percent of                                       ipate in a highly structured medica-
the third month, 72 percent of those         those who were taking quetiapine (79                                   tion administration and tracking
taking conventional agents (6,258 of         of 242) (χ2=46.98, df=1, p<.001).                                      process. To receive clozapine, pa-
8,729 patients) were persistent in                                                                                  tients were required to make at least
therapy compared with 70 percent of          Discussion                                                             eight visits a month to the pharmacy
those taking atypical agents (10,279 of      Our analysis of pharmacy refill                                        and laboratory during the first six
14,763) (χ2=11.27, df=1, p<.001). At         records did not show higher levels of                                  months of therapy. Additional con-
the start of the sixth month, the re-        medication adherence for atypical                                      tacts with a psychiatrist would also
spective rates were 61 percent (3,684        agents as a class than for convention-                                 have been made.
of 6,074) and 55 percent (5,466 of           al agents. This finding runs counter to                                   Moreover, this process demanded
9,922) (χ2=47.77, df=1, p<.001). At          the presumed relationship between                                      heightened accountability on the part
the start of the ninth month, the rates      the improved side effect profile of                                    of psychiatrists and pharmacists, who
were 48 percent (820 of 1,723) and 44        atypical agents and long-term adher-                                   were responsible for certifying week-
percent (1,242 of 2,823) (χ2=5.58,           ence (2). Notably, the risk of patients’                               ly or bimonthly white blood cell
df=1, p<.02). Persistence rates were         discontinuing medication was not                                       counts and reporting results of blood
compared for the different drugs             uniform across points in therapy.                                      tests to a registry for four weeks after
within the group taking conventional            Our data also suggest a risk profile                                a patient discontinued clozapine.
agents and the group taking atypical         for discontinuation of antipsychotic                                   This process may have resulted in
agents. Within-group rates were sim-         medication that is greatest at the                                     better patient engagement in therapy
ilar for the drugs in each group.            start of therapy for all agents. Im-                                   and better identification of patients
   Patients who were taking clozapine        proved follow-up of patients during                                    who missed scheduled blood draws
were significantly more likely to con-       the first 30 to 45 days of therapy or                                  or medication. However, a selection
tinue medication therapy than pa-            after discharge from inpatient care                                    bias for more compliant patients or
tients who were taking conventional          may help reduce the risk of discon-                                    for patients who have better social or
agents and patients who were taking          tinuation.                                                             systems support may have con-
atypicals other than clozapine (Figure          The high and sustained rate of                                      tributed to the high level of persist-
1). Of the patients who were taking an       medication adherence in this study                                     ence demonstrated by patients taking
atypical agent, 71 percent of those          among patients taking clozapine dif-                                   clozapine.
who were taking clozapine (81 out of         fers from rates found in studies that                                     The extent of poor adherence
114 patients) persisted with therapy         have suggested that poor medication                                    among patients taking antipsychotic
at the start of the ninth month, com-        adherence among patients taking an-                                    medication observed in this study is
pared with 40 percent of those who           tipsychotics is the almost unavoidable                                 similar to that in previous reports.
PSYCHIATRIC SERVICES   o September 2001 Vol. 52 No. 9                                                                                                   1249
Cramer and Rosenheck (4) found             pharmacists may have entered incor-        Acknowledgment
that the adherence of patients taking      rect information into the system or        The authors thank Marc Glassman, Ph.D.,
medications for psychosis and mood         patients may have changed residence,       for assistance with the data analysis.
disorders differed little from that of     obtained medication refills through
patients taking medication for physi-      other pharmacy chains, or been re-         References
cal disorders. Teaching patients to        hospitalized.                              1. American Psychiatric Association: Practice
use a “feedback” method that links                                                       guidelines for the treatment of patients
                                                                                         with schizophrenia. American Journal of
daily activities with the process of       Conclusions                                   Psychiatry 154(suppl 4):1–63, 1997
taking medication is one approach          Our findings suggest that the im-
                                                                                      2. Norquist G, Hyman SE: Advances in un-
that can be used to increase rates of      proved side effect profile of atypical        derstanding and treating mental illness: im-
adherence (5,6). Other methods of          agents may not automatically ensure           plications for policy. Health Affairs 18(5):
increasing adherence, such as mail-        high levels of persistence with med-          32–47, 1999
ing reminders to patients at strategic     ication therapy, as is commonly as-        3. Fenton WS, Blyler CR, Heinssen RK: De-
points in therapy, are also being in-      sumed. The high rate of persistence           terminants of medication compliance in
                                                                                         schizophrenia: empirical and clinical find-
vestigated.                                among patients taking clozapine sug-          ings. Schizophrenia Bulletin 23:637–651,
   A limitation of our study is the lack   gests that premature medication dis-          1997
of collateral diagnostic information.      continuation may be not just a patient     4. Cramer JA, Rosenheck R: Compliance with
Persistence with atypical agents was       issue, but a larger systems issue that        medication regimens for psychiatric disor-
likely underestimated in our data set,     may be responsive to processes that           ders. Psychiatric Services 49:196–201, 1998
because for some patients short-term       improve medication administration          5. Cramer JA: Microelectronic systems for
use of atypical agents may have been       and patient tracking. Given the hu-           monitoring and enhancing patient compli-
legitimate; for example, atypical an-      man and financial costs of poor ad-           ance with medication regimens. Drugs
                                                                                         49:321–327, 1995
tipsychotics are used for affective dis-   herence to antipsychotic medication,
orders and dementia. In addition,          population-based processes for im-         6. Cramer JA, Rosenheck R: Enhancing med-
                                                                                         ication compliance for people with serious
some limitations are associated with a     proving medication adherence are              mental illness. Journal of Nervous and
pharmacy database; for example,            ripe for investigation. o                     Mental Disease 187:52–54, 1999

                     Institute on Psychiatric Services Scheduled for October
                     The Institute on Psychiatric Services—the American Psychiatric Association’s an-
                     nual conference on clinical care and service delivery issues—will be held Octo-
                     ber 10–14 at the Renaissance Orlando Resort in Orlando, Florida. Harvey Blue-
                     stone, M.D., of New York City, is chairperson of the 2001 institute scientific pro-
                     gram committee.
                       A day-by-day overview of the institute program appeared in the June issue of
                     Psychiatric Services. The overview is also available on APA’s Web site at
            Order a copy by calling APA’s Answer Center at 888-357-7924.
                     For additional information, contact Jill Gruber, associate director, Institute on
                     Psychiatric Services, APA, 1400 K Street, N.W., Washington, D.C. 20005; tele-
                     phone, 202-682-6314; e-mail,

1250                                                                      PSYCHIATRIC SERVICES   o September 2001 Vol. 52 No. 9