HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Garrard, J. Articles by Leppik, I. Search for Related Content PubMed PubMed Citation Articles by Garrard, J. Articles by Leppik, I.

Factors Associated With Antiepileptic Drug Use Among Elderly Nursing Home Residents

^a Division of Health Services Research and Policy, School of Public Health,
^b Division of Experimental and Clinical Pharmacology, College of Pharmacy, University of Minnesota, Minneapolis
^c Pharmacy Corporation of America, PharMerica, Fridley, Minnesota
^d MINCEP Epilepsy Care, Minneapolis, Minnesota

Judith Garrard, Division of Health Services Research and Policy, School of Public Health, University of Minnesota, 420 Delaware Street, SE, Minneapolis, MN 55455-0381 E-mail: jgarrard{at}tc.umn.edu.

William B. Ershler, MD

	Abstract

Top Abstract Methods Results Discussion Appendix References

 
Background. Epilepsy, a chronic condition defined as two or more recurrent, unprovoked seizures, has the highest incidence at the end of life. Antiepileptic drugs (AEDs) are the primary therapeutic mode. Approximately 10%–11% of elderly nursing home residents receive one or more AEDs, a higher prevalence than would be expected in this age group. In the research literature, there is not a clear explanation of variations in AED use in nursing homes. The purpose of this study was to examine the prevalence and variations in use of AEDs by resident characteristics, AEDs used, drug dosage, and AED combinations in treatment regimens.

Methods. This was a retrospective, cross-sectional study of residents in a convenience sample of nursing homes in 24 states and the District of Columbia. The unit of analysis was the individual resident. The study period was a single day in 1995. Bivariate and multivariate analyses were used to test differences.

Results. The prevalence of AED use was 10.5% across all elderly residents. In a multivariate analysis, factors associated with AED treatment included seizure indication, age group, and geographic region. AED use by age group showed declining use as the residents aged, from 65–74 to 75–84 to 85 years.

Conclusions. The inverse relationship between AED use and age group was unexpected because the incidence of epilepsy increases with advancing age. This finding raises important questions about the future use of these drugs in elderly nursing home residents.

EPILEPSY, a chronic condition defined as two or more recurrent, unprovoked seizures, affects approximately 2.3 million Americans and has an annual cost of $12.5 billion ⁽¹⁾. Historically, the onset of epilepsy was believed to occur primarily in childhood; however, recent studies have shown that incidence rates actually form a U-shaped curve, with the highest rates at the end of life. In the United States, the estimated incidence rate for epilepsy in infants up to 1 year of age is 82 per 100,000 compared with 139 for people aged 75 years and older ⁽²⁾.

The primary cause of epilepsy in older people is stroke, which accounts for 30%–40% of all cases ⁽³⁾. The prevalence of epilepsy varies by age group, from 0.68% for people 65–74 years to 1.40% for those 75 years and older ⁽⁴⁾. The prevalence among nursing home residents aged 65 and older ranges from 5% to 9% ⁽⁵⁾⁽⁶⁾.

Antiepileptic drugs (AEDs) are the primary treatment for seizures associated with epilepsy. Although the distribution of AEDs across the lifespan has not been reported in the literature, the prevalence of the condition, together with the increasing numbers of elderly adults in the population, suggests that the number of people 65 years and older likely to receive an AED will increase substantially in the next several decades ⁽⁷⁾. Adding to this rapid growth rate is the expanding use of AEDs for other conditions, such as agitation, mania, neuropathic pain, and tremor, which are common among older people ^(8)(9)(10).

The pharmacokinetics of the most frequently prescribed AEDs are complex, which makes dosing and monitoring difficult in any age group; however, treatment decisions become more complicated with the older patient ⁽¹¹⁾. Age-related alterations in physiology can affect AED pharmacology and limit the choice of medications ⁽¹²⁾. There is a lack of safety and efficacy information to guide clinical decisions about AED therapy in older people ⁽¹³⁾. Finally, the complexity of medical problems and comedications among older people requiring AEDs can result in an increased likelihood of drug-related problems and drug interactions, which in turn can affect seizure control and toxicity ^(14)(15)(16).

These problems are exacerbated among nursing home residents who are likely to be at greatest risk for clinical complications and frailty. The prevalence of AED use in the nursing home population varies between 10% and 11% ^(6)(17). Studies using bivariate analyses have shown that men and persons 65–84 years have greater use of AEDs than do women and persons 85 years or older, respectively ⁽¹⁷⁾. No multivariate studies have been reported that adjust simultaneously for gender, age, and condition, however.

Of the AEDs available on the market, the medications most likely to be used by elderly nursing home residents were (in descending order) phenytoin, carbamazepine, phenobarbital, and clonazepam ⁽¹⁷⁾. In another study, valproic acid rather than clonazepam ranked among the four most frequently prescribed AEDs, a pattern consistent with AED use in younger adult populations ⁽⁶⁾. More than 80% of the residents in both studies used a single AED (AED monotherapy). Of those who used two or more AEDs (AED polytherapy), the specific medications and frequency of use varied widely.

The lack of information about AED use in nursing homes makes it difficult to propose rational therapy guidelines, assess the appropriateness of drug selection and dosing, evaluate therapeutic outcomes, and track trends in drug use ⁽¹⁸⁾. There is considerable variability in indications, use of specific medications, dosage, drug combinations, and the characteristics of patients receiving an AED. These variations could be explained by differences in study design, methods of analysis, or clinical practice. Practice variations could be associated with regional differences in treatment decisions, although little research on this issue has been reported to date.

The goal of this study was to systematically examine the prevalence of AED therapy among elderly nursing home residents and examine variations in use by resident characteristics, specific AEDs used, drug dosage, and AED combinations in treatment regimens. For purposes of this study we limited our analysis to the four AEDs most commonly prescribed in adults: carbamazepine, phenobarbital, phenytoin, and valproic acid. The research questions in this study were threefold:

Factors associated with any AED treatment.—What is the association between AED treatment and characteristics of nursing home residents by age group, gender, seizure indication, population density (urban or rural), and geographic region?

Factors associated with each AED.—What is the association between use of each of four major AEDs among nursing home residents by the same resident characteristics (i.e., age group, gender, seizure indication, and population density and geographic region of the nursing home)? The four AEDs included in this study are carbamazepine, phenobarbital, phenytoin, and valproic acid.

Variations in AED daily dose.—Among nursing home residents who received an AED, what factors were associated with variations in daily dose of the medication for each of the four major AEDs?

	Methods

Top Abstract Methods Results Discussion Appendix References

 
Study Design
This was a retrospective, cross-sectional study of nursing home residents in 24 states and the District of Columbia. All of the nursing homes in this convenience sample had a contract with Computran, a provider of electronic medical record services to long-term care facilities.

The unit of analysis was the individual resident. The study period was a single day during a 30-day period between May and June 1995; the study day varied by nursing home within this 1-month period.

This project included a substudy of the accuracy of the Computran electronic data compared with source data in a subset of seven nursing homes in the Minneapolis–St. Paul metropolitan area. The results of the substudy are described in the Methods section.

Subjects
Subjects were all residents, 65 years or older, living in 346 nursing homes on the study day. Missing data were less than 1% for geographic region, 1% for gender, and 3% each for seizure indication and population density. There was no imputation of missing data in the analysis.

Data Source and Variables
Data source..-- Data were obtained from Computran electronic medical records for epilepsy or seizure indication, medication orders, and patient characteristics. Information about name of drug, dosage, route of administration, and administration status (scheduled or prn) of AED therapy was available, and both brand and generic names were used in identifying these medications.

Medical records data in the Computran system are updated monthly by nursing home personnel, and pharmacy data are input daily by pharmacists as part of each medication order. Geographic information about each nursing home was available from zip codes.

Dependent variables: AED treatment..-- AED treatment was defined as a prescription medication administered on a scheduled or routine basis during the 1-day study period; prn medications were excluded. The dependent variable for the first research question concerning factors associated with any AED treatment was presence or absence of the treatment. All AEDs available on the market during the study period were included, whether they were among the most commonly used AEDs or not ⁽¹⁹⁾. AEDs used by one or more residents in this sample are listed in the appendix.

In the analysis for the second research question, concerning factors associated with type of drug, each of the four most commonly used AEDs (phenytoin, carbamazepine, valproic acid, and phenobarbital) was used separately as a dependent variable. This variable was defined as presence or absence of the AED among all residents who took one or more of any AED.

Dose of each of the four AEDs was the dependent variable in the third question. In constructing this dependent variable, total milligrams per day of all formulations used for each AED were summed for a total daily dose equivalent per resident. All calculations were standardized to milligrams and daily administrations. When phenytoin is administered as a suspension, the actual dosage is slightly increased in milligrams due to a greater drug concentration; therefore, the calculation for phenytoin suspension was adjusted by multiplying the daily dose by a factor of 1.09.

Main effects..-- For all three research questions, the main effects were gender, age group, seizure indication, rural or urban location of nursing home, and geographic region within the continental United States. AED polytherapy was an additional main effect in the third question concerned with daily dose. These variables were operationally defined as follows.

Gender and age..-- Information about age and gender was available in the electronic record. Age in years was converted to the standard age groups on the study day: 65–74 years (young-old), 75–84 years (old), and 85 years (oldest-old) for the multivariate analysis to address the first two questions about use and as a continuous variable for the third question about dosage.

Seizure indication..-- A seizure indication in the electronic record was defined as epilepsy, seizure, or convulsion. Rarely was information about the condition provided in the record as an ICD-9 code, but when it was, the ICD-9 codes 178 and 345 were included in this definition of seizure indication.

Population density..-- Rural or urban designation was based on the Standard Metropolitan Statistical Area definitions as defined by the U.S. Census Bureau ⁽⁷⁾. Zip code of the nursing home address and the 1990 census were used to determine rural or urban status of the nursing home.

Geographic region..-- To examine the possibility of geographic differences, each nursing home was assigned to one of four non-overlapping geographic regions as defined by the Census Bureau ⁽²⁰⁾. The states included in each region and characteristics of the sample by gender and age group are shown in Table 1 .

Substudy of Agreement Between Electronic and Source Data
Agreement between the Computran electronic data and the on-site source data in the residents' records was examined in a substudy in seven nursing homes in the Twin Cities metropolitan area. Records of all residents, 65 years and older, with current AED orders (n = 42) in these seven facilities were identified in the Computran database. The gold standards for this substudy consisted of (i) information in nursing home records for diagnostic information and (ii) AED use by residents as documented in the medication administration record (MAR) located at each nursing home.

All nonelectronic records (Minimum Data Set, Resident Assessment Form, Physician Order Form, admission records to the facility and hospital, and discharge records) were included in the search for a seizure indication or epilepsy diagnosis. Medicare-certified nursing homes are required by federal regulation to maintain an MAR that is a daily record of all prescription and nonprescription medications administered to nursing home residents. Only certified nursing home personnel are permitted to administer medications, and the information recorded for each resident in the MAR must include the name of the medication, dosage, date and time of administration, and the initials of the person who administered it.

Agreement was defined as the percentage of agreement between the electronic record and the nursing home record (for diagnostic information) or MAR (for AED medications) for each of three variables: (i) presence or absence of a diagnosis of epilepsy or seizure disorder; (ii) presence or absence of AED medication orders during the study period; and (iii) exact agreement in milligrams by specific AED. Kappa, a chance-corrected statistical measure of agreement, was also computed for each of the three variables ⁽²¹⁾.

Results show that the agreement between the nursing home record and the electronic record was 86% for diagnostic information and 100% for presence or absence of AED medications Exact agreement between dosage administered (from the MAR) and dosage ordered (electronic record) ranged from 79% to 100%, depending on the AED. Kappa values ranged from .81 to 1.00. Results of this substudy suggest a strong level of agreement for each of the three variables between the electronic record and each respective gold standard. Kappa values greater than .75 indicate excellent agreement beyond chance ⁽²²⁾.

Statistical Analysis
Bivariate and multivariate analyses were completed for all of the research questions. The bivariate analysis consisted of either a chi-square or a one-way analysis of variance, depending on the measurement scale of the dependent variable.

The same three-step procedure for the multivariate analysis of the first two research questions consisted of the following. Initially, a logistic regression model was computed that included only main effects. A second logistic regression model was run, based on all statistically significant main effects and their pairwise interactions. A final logistic regression model was computed that included significant main effects and significant interaction terms. A hierarchical rule was followed: a main effect (statistically significant or not) was included if the effect was part of a statistically significant interaction term. Only the results of the third model are described in the Results section.

The analysis for the third research question concerning factors associated with AED dose consisted of ordinary least squares regression, because the dependent variable was continuous. This regression analysis followed the same three-step procedure as described previously for the logistic regressions.

The possibilities of confounding and heteroscedasticity were examined in the final models. Neither was statistically significant in any of the analyses.

	Results

Top Abstract Methods Results Discussion Appendix References

 
Subject and Treatment Characteristics
Subject profile..-- The mean age of subjects was 83.78 years (standard deviation, 8.13). Their distribution by gender was 76% female and by age group, 15%, 36%, and 49% for ages 65–74 years, 75–84 years, and 85 years, respectively. This distribution is similar to that of a nationally representative sample of nursing home residents ⁽²³⁾. Subject characteristics are summarized in Table 1 and Table 2 .

Proportionately more males than females . Age group and AED prevalence had an inverse relationship . Among residents in the 65–74 year age group, 24% were taking an AED, compared with 12% among those aged 75–84 years, and 6% of those 85 and older. A significantly greater proportion of residents with a seizure indication received an AED than those without such an indication (; Fig. 1) These results show that for each age group, two-thirds of the residents using one or more AEDs on the study day had documentation of a seizure indication in the electronic record. Residents living in nursing homes in urban locations had greater use of AEDs (11.1%) than those in rural areas (9.5%) .

View larger version (28K): [in this window] [in a new window]   Figure 1. Percentage of residents receiving an antiepileptic drug (AED) by seizure (Sz) indication within age groups.

Residents living in Southern states had significantly lower AED prevalence than those living in the Northeast. There were no other geographic differences in AED treatment rates. The significant gender and rural–urban differences seen in the bivariate analysis became nonsignificant in the multivariate analysis.

Estimated probabilities of AED treatment by subgroups..-- An alternative approach to examining multivariate results is through the use of estimated probabilities based on statistically significant main effects. In other words, after adjusting for significant differences in the logistic regression model, what is the estimated likelihood of a nursing home resident having AED use in each of the subgroups? The estimated probabilities calculated from exponentiation of the beta values in the logistic regression analysis are shown in Table 6 ⁽²⁴⁾.

Factors Associated With Specific AED Medications
Odds ratios (OR) resulting from the logistic regression analysis of each of the AEDs (phenytoin, carbamazepine, phenobarbital, and valproic acid) are summarized in Table 7 . The significant effects varied from drug to drug.

Carbamazepine..-- Two of the five variables were significantly associated with prevalence of carbamazepine: age group (OR 0.53, p .0001 for 75–84 years; OR 0.27, p .0001, for 85 years; both compared with residents 65–74 years) and seizure indication (OR 9.32, p .0001 compared with those without seizure indication). The direction of the differences was the same as for any AED use and for phenytoin: the younger the resident, the more likely he or she was to use carbamazepine, and a seizure indication was strongly and significantly associated with carbamazepine use.

Phenobarbital..-- Two of the five effects were significant in the use of phenobarbital. Presence of a seizure indication was strongly associated with the use of this AED. The odds ratio was 66.57 (p .0001 compared with those without a seizure indication). Use of phenobarbital also differed by geographic region. Each of the geographic regions South, Midwest, and West differed from the Northeast, but none of those three differed from one another.

Valproic acid..-- Use of valproic acid was significantly associated with three of the five variables: age group (OR 0.39, p .01, for 75–84 years; OR 0.19, p .0001, for 85 years; both compared with residents 65–74 years), seizure indication (OR 4.37, p .0001, compared with those without seizure indication), and geographic region. Similar to the results for any AED, use of valproic acid showed a statistically significant difference between the Northeast and the South (OR 0.46, p .01) but not between the reference group or either of the other two geographic regions; differences between the South and the other two geographic regions were also statistically significant. The younger the nursing home resident, the more likely that he or she would receive valproic acid, especially if there was documentation of a seizure indication; however, residents living in the South were statistically less likely to use this medication than those in any other region in the United States.

Factors Associated with Variations in AED Daily Dose
In the third research question, we were interested in examining the association between the predictor variables and standardized daily dose (in milligrams) of each of the four AEDs among residents who took one or more AEDs. The number of subjects in each analysis varied: phenytoin . Because the dependent variable was continuous, we used a multiple regression model for each analysis. For the predictor variables, age (in years) was used as a continuous variable and AED polytherapy (yes/no) was added as a potential main effect, in addition to the other variables (gender, seizure indication, rural or urban, and geographic region).

The results, summarized in Table 8 , show that one or more of these six variables were statistically associated with three of the AEDs: phenytoin, carbamazepine, and valproic acid. None of these six variables was associated with a daily dose of phenobarbital.

Carbamazepine..-- Dose of carbamazepine was associated with three variables: age beta - , geographic region (beta -, and AED polytherapy beta . Thus, the daily dose of carbamazepine decreases as the person becomes older. Those in the South have a lower daily dose of carbamazepine by a daily average of 96 mg compared with residents in the Northeast. Residents with two or more AEDs (AED polytherapy) have a higher dose of carbamazepine, by an average of 135 mg, compared with those who use only carbamazepine. Although these three main effects were statistically significant, the amount of variance accounted for by this model was only 11%. Thus, other variables need to be considered in understanding what factors are associated with carbamazepine dose.

Valproic acid..-- For valproic acid, daily dose was associated with four of the six variables: age beta -beta beta beta . The amount of variance accounted for was 26%, a moderate but not strong association. In general, these results are consistent with the other two AEDs in showing an inverse relationship with age, higher dosage in the West than in the Northeast, and higher dosage for those with AED polytherapy. In none of the other AEDs was dose associated with seizure indication; however, for valproic acid, the resident with a seizure indication had a higher dose, by 238 mg on the average, compared with his or her counterpart who did not have such an indication in the record.

	Discussion

Top Abstract Methods Results Discussion Appendix References

 
The prevalence of AED therapy over a 1-day study period was 10.5% in this convenience sample of more than 21,000 nursing home residents aged 65 years or older. Extrapolation of these findings suggest that as many as 162,600 people were likely to be receiving an AED among the 1.5 million elderly residents in U.S. nursing homes in 1995 ⁽²⁵⁾.

Three of five factors in this multivariate analysis were significantly associated with AED use: age group, seizure indication, and geographic region. The inverse relationship between AED use and age group was unexpected. Elderly nursing home residents are more likely than people in the community to have experienced a stroke, which is a significant risk factor for the development of epilepsy. It can be reasoned that the percentage of those with a seizure indication and AED treatment would increase by age group. Additional AED use as treatment for other neuropsychiatric disorders might also be expected to increase with age. These circumstances predict that a positive relationship should exist between AED use and advancing age. Nonetheless, we found both a declining seizure indication and AED prevalence by age group.

There are several possible explanations for these findings: (i) a healthy survivor effect in which those with an epileptic condition died at earlier ages (e.g., 65–74 years) and more of those without the condition or need for AED treatment survived into their eighties or longer; (ii) residents in the younger age group (65–74) may have been admitted with more acute conditions associated with either the condition or the need for AED treatment ⁽²⁶⁾; (iii) underdocumentation of the condition or AED treatment; (iv) underdiagnosis of the condition and therefore lack of treatment; or (v) accurate diagnosis but undertreatment. The third and fourth explanations seem less likely, because a seizure can be a dramatic event; however, a history of seizure may be underdocumented. Both undertreatment of epilepsy with advancing age and a survivor effect are plausible explanations; however, longitudinal studies are needed to determine the underlying reason for this finding.

In this study, approximately one out of four residents in the 65–74 year age group took an AED, which places this drug category among the most frequently prescribed class of neuropsychiatric medications in nursing homes ⁽²⁷⁾. This has important pharmacoepidemiologic and pharmacoeconomic implications. Given the expected growth in the 65–74 year age group over the next several decades, the number of people receiving AEDs can be expected to increase substantially. The continued use of the inexpensive AEDs most commonly prescribed, which have serious side effects and interact with other medications, will greatly increase the risk of adverse events. In contrast, a switch to more easily managed but expensive newer AEDs will add substantially to medication costs, but potentially reduce the risk and cost of subsequent drug-related problems ⁽²⁸⁾. Because many of the newer AEDs became available on the market after 1995, when this study was conducted, an estimate of the proportion of elderly nursing home residents with newer versus older AEDs is not available.

As would be expected, clinical indication was strongly associated with use of an AED; 79% of all residents with a seizure indication had routine use of an AED. (Of those with one or more AEDs, 67% had an epilepsy or seizure indication.) Prevalence of AED use adjusted for age and seizure indication was highest in nursing homes in the Northeast and lowest in the South; however, the prevalence did not differ between the Northeast and either of the other two geographic regions, the Midwest and West. There were no rural–urban differences in AED prevalence.

Although a gender difference in AED use has been reported in bivariate analyses ⁽¹⁷⁾, this factor was not significant in the multivariate model, probably because gender and age are confounded. Eighty-three percent of the residents 85 and older were women, a typical age distribution in U.S. nursing homes.

Of the four major AEDs, phenytoin was the most commonly used (6%), followed by carbamazepine (2%), phenobarbital (2%), and valproic acid (1%). Fourteen percent of the residents were treated with two or more AEDs; the most frequent combination was phenytoin and phenobarbital.

Multivariate models of each of the four AEDs showed that use of phenytoin was associated with presence of seizure indication, an inverse relationship by age group, and differences by geographic region, with greater use in the Midwest and West than in the Northeast. Carbamazepine showed a similar pattern with two of the five factors, seizure indication and age group. Valproic acid was associated with the same three factors as phenytoin; however, there was significantly less use of valproic acid in the South than in any of the other three regions.

In the multivariate models for each of the four AEDs, seizure indication was consistently the strongest factor associated with use of each AED. Age group was also a statistically significant factor in three of the four AEDs (phenytoin, carbamazepine, and valproic acid). The pattern of use by age group was consistent across the three AEDs, showing declining use with age. Residents in the three age groups did not differ statistically in their use of phenobarbital in the multivariate model.

AED use differed by geographic region; however, the patterns were not consistent. Phenytoin was more likely to be used by residents living in nursing homes in the Northeast and South; valproic acid was in greater use in the Northeast, Midwest, and West than in the South; and phenobarbital was in greater use in the Northeast than in the other three regions. The four geographic regions did not differ in use of carbamazepine.

The mean daily dosages for carbamazepine and valproic acid were lower, whereas phenobarbital and phenytoin dosages were comparable to those used in younger adults ⁽¹²⁾. Dosage of phenobarbital was not associated with any of the six factors examined in the multivariate analysis: age, seizure indication, geographic region, AED polytherapy, gender, or rural–urban location. In the regression models for the other three drugs, age was consistently and inversely related to dosage.

The interpretive picture that emerges is one in which the older the person, the less likely he or she would receive any AED. If an AED is taken, then the dosage is lower with advancing age. Whether this was due to declining need for treatment because of lower prevalence of seizure indication or caution by the clinician cannot be determined from this dataset.

Valproic acid was the only AED in this analysis in which seizure indication was significantly associated with a higher daily dose (238 mg on the average). There were geographic differences in dosage, but the patterns differed by drug. Phenytoin had higher average dosage (21 mg) in the West than in the Northeast, carbamazepine had a lower average dosage (96 mg) in the South than in the Northeast, and valproic acid had a higher dosage level (158 mg) in the West than in the Northeast. These differences suggest the possibility of area variations that need further explanation.

AED polytherapy was associated with higher doses in two of the three models. When a second or third AED was used, the dose of carbamazepine increased by 135 mg, on the average, and valproic acid increased by 492 mg. Higher doses with AED polytherapy could occur because the epilepsy is difficult to treat, which may be the reason why a second or third AED was added. Alternatively, the higher doses may compensate for the more rapid drug elimination resulting from drug interactions ⁽²⁹⁾.

The robustness of the models of daily dose differed. The valproic acid model, with four factors, accounted for 26% of the variance, compared with 1% of the variance in the phenytoin model with two factors and 11% of the variance in the carbamazepine model with three factors. None of the six factors used to model phenobarbital captured variations in dosage. Based on the variance in carbamazepine dosages accounted for by the model, the three factors for this agent are probably not sufficient for predictions.

Although this dataset had a large number of subjects, their geographic distribution was limited to 24 states and the District of Columbia. Thus, these results may not be representative of nursing home residents throughout the United States. AED use over a 1-day study period is also problematic because variation may be greater over a longer time period, and seasonality may be a factor. An improved study design would span 12 months. Another problem was the lack of information about actual diagnosis and severity of the epileptic condition. A seizure indication suggests that the condition may exist, but a clearly stated diagnosis would be preferable. In general, however, these results provide an initial overview of the prevalence of AED use and the likelihood of association among demographic variables, drug use, and dosage. The conclusions that can be drawn from these results suggest that the prevalence of AED use is considerably higher in nursing homes (10%) than among elderly people living in the community, and in particular, a high percentage (25%) of young-old nursing home residents are likely to receive an AED.

	Acknowledgments

 
This study was supported by grants from Novartis Pharmaceuticals Corporation and the National Institutes of Health/National Institute of Neurological Disorders and Stroke (Grant P50-NS16308). The assistance of John Rarick and Xiaohong Chen, both at the University of Minnesota, and Trevor Nolte, who at the time of the study was with Pharmacy Corporation of America, is gratefully acknowledged.

	Footnotes

 
Thomas Lackner is now in the Division of Experimental and Clinical Pharmacology, College of Pharmacy, University of Minnesota, Minneapolis. Nina Graves is now at Medtronic, Inc, Minneapolis, MN.

Received July 1, 1999

Accepted November 8, 1999

	Appendix

Top Abstract Methods Results Discussion Appendix References

 
Antiepileptic Drugs Used by One or More Residents in This Study
Phenytoin

Dilantin–125 suspension

Dilantin

phenytoin sodium

Carbamazepine

carbamazepine

Tegretol

Valproic acid and/or sodium valproate

Depakene

Depakote

divalproex

valproate

valproic acid

Phenobarbital

Mebaral

phenobarbital

Primidone

Mysoline

primidone

Other

Celontin

clonazepam

Felbatol

Klonopin

Mesantoin

	References

Top Abstract Methods Results Discussion Appendix References

 

1. Epilepsy Foundation of America. Epilepsy: A Report to the Nation. Available at http://www.efa.org/news/nationa/nation.html#1. Accessed April 21, 1999.
2. Hauser WA, Annegers JF, Kurland LT, 1993. Incidence of epilepsy and unprovoked seizures in Rochester, Minnesota: 1935–1984. Epilepsia. 34:453-468. [Medline]
3. Hauser WA, 1997. Epidemiology of seizures and epilepsy in the elderly. Rowan AJ, Ramsay RE, , ed.Seizures and Epilepsy in the Elderly 7-20. Butterworth-Heinemann, Boston, MA.
4. Hauser WA, Annegers JF, Kurland LT, 1991. Prevalence of epilepsy in Rochester, Minnesota: 1940–1980. Epilepsia. 32:429-445. [Medline]
5. Hauser WA, 1998. Epidemiology of seizures and epilepsy in the elderly. Willmore LJ, , ed.Epilepsy in the Elderly—Current Concepts in Understanding and Treatment American Epilepsy Society, Hartford, CT.
6. Schacter SS, Cramer GW, Thompson GD, Chaponis RJ, Mendelson MA, Lawhorne L, 1998. An evaluation of antiepileptic drug therapy in nursing facilities. J Am Geriatr Soc. 46:1137-1141. [Medline]
7. 1990 Census of Population and Housing. Summary Tape, File 3B [CD-ROM]. Washington, DC: US Census Bureau; 1993.
8. Grossman F, 1998. A review of anticonvulsants in treating agitated demented elderly patients. Pharmacotherapy. 18:600-606. [Medline]
9. Ferrell BA, 1991. Pain management in elderly people. J Am Geriatric Soc. 39:64-73. [Medline]
10. Lou JS, Janovic J, 1991. Essential tremor: clinical correlates in 350 patients. Neurology. 41:234-238.
11. Thomson AH, Brodie MS, 1992. Pharmacokinetic optimisation of anticonvulsant therapy. Clin Pharmacokinet. 23:216-230. [Medline]
12. Cloyd JC, 1997. Commonly used antiepileptics: age-related pharmacokinetics. Rowan AJ, Ramsay RE, , ed.Seizures and Epilepsy in the Elderly 219-228. Butterworth-Heinemann, Boston, MA.
13. Ramsay RE, 1997. Response to treatment and adverse drug effects in the elderly. Rowan AJ, Ramsay RE, , ed.Seizures and Epilepsy in the Elderly 59-64. Butterworth-Heinemann, Boston, MA.
14. Cloyd JC, Lackner TE, Leppik IE, 1994. Antiepileptics in the elderly. Pharmacoepidemiology and pharmacokinetics. Arch Fam Med. 3:589-598. [Abstract/Free Full Text]
15. Willmore LJ, 1996. Management of epilepsy in the elderly. Epilepsia. 37: (suppl 6) S23-S33.
16. Fouts M, Hanlon J, Pieper C, Perfetto E, Feinberg J, 1997. Identification of elderly nursing home facility residents at high risk for drug-related problems. Consult Pharm. 12:1103-1111.
17. Lackner TE, Cloyd JC, Thomas LW, Leppik IE, 1998. Antiepileptic drug use in nursing home residents: effect of age, gender, and comedication on patterns of use. Epilepsia. 39:1083-1087. [Medline]
18. Sirven JI, 1998. Epilepsy in older adults: causes, consequences and treatment. J Am Geriatr Soc. 46:1291-1301. [Medline]
19. Guide to antiepileptic agents, 1995. Pharm Pract News. 1995;22(2):14–17.
20. US Census Bureau 1994. County and City Data Book, 1994: A Statistical Abstract Supplement 12th ed. US Department of Commerce;, Washington, DC.
21. Fleiss JL, 1981. Statistical Methods for Rates and Proportions 2nd ed. John Wiley & Sons;, New York.
22. Landis JR, Koch GG, 1977. The measurement of observer agreement for categorical data. Biometrics. 33:159-174. [Medline]
23. Hing E, Sekscenski ES, 1987DHHS publication (PHS) 87-1409.. Use of health care—nursing care. Havlik RJ, Liu BM, Kovar MG, , ed.Health Statistics on Older Persons, United States, 1986. Vital and Health Statistics, Series 25, No 25 Public Health Service, US Government Printing Office, Hyattsville, MD.
24. Selvin S, 1995. Goodness-of-fit. Practical Biostatistical Methods 397Duxbury Press, Albany, NY.
25. 1995 National Nursing Home Survey. National Center for Health Statistics, Advanced Data Number 280, January 23, 1997. Washington, DC: US Government Printing Office; 1997.
26. Garrard J, Buchanan JL, Ratner ER, et al. 1993. Differences between nursing home admissions and residents. J Gerontol Soc Sci. 48:S301-S309.
27. Tobias DE, Pulliam CC, 1997. General and psychotherapeutic medication use in 878 nursing facilities: a 1997 national survey. Consult Pharm. 12:1401-1408.
28. Leppik IE, 1997. Choosing antiepileptic drugs for the elderly. Rowan AJ, Ramsay RE, , ed.Seizures and Epilepsy in the Elderly 65-71. Butterworth-Heinemann, Boston, MA.
29. Riva R, Albani F, Contin M, Baruzzi A, 1996. Pharmacokinetic interactions between antiepileptic drugs—clinical considerations. Clin Pharmacokinet. 31:470-493. [Medline]

This article has been cited by other articles:

				J. E. Ahn, J. C. Cloyd, R. C. Brundage, S. E. Marino, J. M. Conway, R. E. Ramsay, J. R. White, L. C. Musib, J. O. Rarick, A. K. Birnbaum, et al. Phenytoin half-life and clearance during maintenance therapy in adults and elderly patients with epilepsy Neurology, July 1, 2008; 71(1): 38 - 43. [Abstract] [Full Text] [PDF]

				A. Birnbaum, N. A. Hardie, I. E. Leppik, J. M. Conway, S. E. Bowers, T. Lackner, and N. M. Graves Variability of total phenytoin serum concentrations within elderly nursing home residents Neurology, February 25, 2003; 60(4): 555 - 559. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Garrard, J. Articles by Leppik, I. Search for Related Content PubMed PubMed Citation Articles by Garrard, J. Articles by Leppik, I.