Physician Preference for Antiepileptic Drug Concentration Testing

Robert J. Baumann, M.D.

Departments of Neurology and Pediatrics, College of Medicine, University of Kentucky

Melody Ryan, Pharm.D.

Division of Pharmacy Practice and Science, College of Pharmacy and

Department of Neurology, College of Medicine, University of Kentucky

Aaron Yelowitz, PhD.

Department of Economics, University of Kentucky

Correspondence to Dr. Baumann, Kentucky Clinic L445, University of Kentucky, Lexington, Ky 40536-0284

e-mail: Baumann@uky.edu

Fax: 859 323 5943

Office phone: 859 323 6702 extension 242 (Anna Lee Shipman, secretary) or 250 (Dr. Baumann’s direct line)

Keywords: Attitudes, Anticonvulsants, Children, Epilepsy, Pain, Physicians, Preferences, Saliva, Survey

Running title: Preference for Testing

Abstract:

A four item questionnaire asked USA Child Neurology Society members to value painless anticonvulsant concentration monitoring, whether members had ordered a saliva level (the best established painless method) in the last year and whether such levels were available. Value was quantified by time per patient that the physician would willingly expend to arrange for the test. 544/945 questionnaires were returned (58%). When asked the value of a painless method for children, 286/522 (55%) reported willingness to expend 10 to 30 minutes to arrange the test. 498/522 (95%) would use a painless method if available. When asked the value of an immediate sample at home during a seizure or adverse event - a substantial majority, 370/526 (70%), would make an important donation of their own time to arrange for the sample. Only 5% would not utilize it. Just 2/544 respondents had obtained a painless (saliva) concentration while merely 33/544 (6%) perceived such tests as being available. We conclude that child neurologists put a high value on painless anticonvulsant monitoring. These data suggest that a painless method of measuring anticonvulsant concentrations - especially if it could be done at home - would fulfill an unmet need in the care of children with epilepsy.

Introduction

Successful treatment of epilepsy often requires the monitoring of anticonvulsant drug concentrations. Within the usual therapeutic ranges for these drugs, most children have good seizure control with minimum adverse effects.{Hauser & Hesdorffer 2001 2443 /id} Monitoring anticonvulsant concentrations is also useful in evaluating patient adherence to the treatment plan {Leppik 1998 944 /id} and for monitoring the variations induced by changes in weight and metabolism as children grow.{Birnbaum, Kriel, et al. 2001 2444 /id} The review articles that guide American physicians mention using other tissues to monitor anticonvulsant concentrations but virtually all guidance and usual therapeutic values are given for blood and thus require a venipuncture.{Kriel, Birnbaum, et al. 1999 2461 /id}{Leppik 1998 944 /id}{Menkes & Sankar 2002 2462 /id} This is in part because of good evidence correlating the effects of the commonly used anticonvulsants with serum concentrations. Interestingly, there is also good evidence correlating the effects of the commonly used anticonvulsants, phenobarbital, phenytoin and carbamazepine with concentrations as measured in saliva. Saliva is the alternative body fluid, available without causing the patient pain, that is best suited for therapeutic drug monitoring.{Pichini, Altieri, et al. 1996 2460 /id} Nevertheless neither review articles nor discussions with colleagues suggest that saliva levels are commonly utilized in the United States.{Leppik 1998 944 /id}{Birnbaum, Kriel, et al. 2001 2444 /id}

Monitoring with serum anticonvulsant concentrations involves some obvious disadvantages. There is the discomfort and fear associated with the venipuncture necessary to obtain serum. Additionally there is the cost and inconvenience of going to a clinic or hospital to have the blood drawn. The problem of discomfort is magnified with children whose age or limited intellectual ability makes it impossible for them to understand why this pain is “for their own good”. Moreover, parents may not want to subject a child to this discomfort and thus neglect having their child tested and then miss subsequent clinic appointments.{Rylance & Moreland 1981 258 /id}

None of these disadvantages occur with saliva monitoring. There is no pain and no special skill is needed to obtain the sample.{Bailey, Klein, et al. 1997 936 /id}

We have been puzzled by the apparent lack of use of a painless method of therapeutic monitoring. One hypothesis is that the adoption of new technology is dependent upon physician attitudes and behavior. Thus, if saliva monitoring is not being used, this hypothesis posits that physicians are uninterested in the advantages presented by saliva testing such as reducing discomfort and by not requesting saliva levels are preventing the adoption of this technology. We investigated these issues by questionnaire, surveying American child neurologists.

Methods:

With permission of the Medical Institutional Review Board, we mailed a questionnaire to each active United States member of the Child Neurology Society. To be an active member, a physician needs to be “certified in Neurology with Special Qualification in Child Neurology by the American Board of Psychiatry and Neurology” or “eligible to take the examination for certification”.{Child Neurology Society 2000 2445 /id} The vast majority of these physicians are also eligible for certification by the American Board of Pediatrics.

We had Medical Institutional Review Board permission to identify respondents and to send two follow-up questionnaires to physicians who did not respond initially. The questionnaire consisted of four questions. In order to encourage responses and to limit the burden on the surveyed physicians, the questionnaire was limited to a single page. The first two questions asked whether respondents had ordered one or more saliva levels in the last year and whether the respondents believed that such levels were available to them. Two additional questions inquired about the perceived value of painless monitoring to their patients in routine practice and at home monitoring in crisis situations. For these last two questions respondents could choose between the four responses listed in the Table.

Excluded from the survey were the investigators and their colleagues and research associates.          

Results

          Of 1006 mailed surveys, 57 were returned by the postal service as undeliverable and four went to members of our research group giving a denominator of 945 subjects. After 3 mailings, 544/945 questionnaires had been returned (58%). We compared responders and non-responders dividing them into the four regions defined by the U.S. Census Bureau{U.S.Census Bureau 2002 2457 /id}. We could find no statistically significant geographic differences between responders and non-responders. Only 2/544 (0.4%) respondents indicated having obtained a saliva level in the last year (question #1) and just 33/544 (6%) indicated that saliva levels were available to them (question #2). 187/544 (34%) replied that no such levels were available to their practice and 312/544 (57%)‘didn’t know’ (question #2). Of the 33 positive respondents, 12 were located in the Midwest, 8 in the South, 7 in the East and 6 in the West with no predominance in any specific cities.

Question three asked: “How valuable to the care of your pediatric patients would be the ability to obtain anticonvulsant levels by a painless method as opposed to serum which requires a venipuncture?” There were 522 usable responses (Table). 286/522 or 55% of responding physicians thought such an innovation valuable enough that they would expend 10 to 30 minutes of their own time to arrange the test. 498/522 or 95% felt they would use a painless method if it were available. Question four asked: “How valuable to the care of selected patients would be the ability to obtain an immediate sample at home for anticonvulsant level determination at the time of a seizure or adverse event - without the delay necessitated by a trip to a laboratory or emergency department.” A substantial majority of the sample, 370/526, 70%, of the physicians would make a significant donation of their own time to arrange for such a test. Only 5% felt it was not of value (Table).

Discussion:

Our survey indicates that child neurologists place a high value on a painless method of monitoring anticonvulsant concentrations (Table). We sought a value judgment from the respondents and avoided the traditional responses “strongly agree”, “”agree”, “neither agree nor disagree”, and “strongly disagree”. The reason we avoided those responses is that the questions involved pain versus no pain. We were concerned that respondents would feel that for social reasons they were obligated to “agree” with any method that reduced pain whatever their true feelings. Money is commonly used in such surveys to measure value - respondents are asked how much they would be willing to pay to achieve a given outcome. In our opinion, because doctors don’t personally finance the care of their patients, this measure would be unrealistic for a physician survey. On the other hand, it is common for physicians to expend their time without additional remuneration in the care of patients. So we asked physicians to what extent they would be willing to spend their time “to obtain anticonvulsant levels by a painless method”(Table). To our surprise 11% of physicians choose the most expensive option, one-half hour of their time. Another 44% offered 10 minutes of time, the second option. These two groups outnumbered the 212 physicians (41%) who would use the test but didn’t feel it warranted an additional time investment.

A number of physicians who choose the third answer, “...I might use the test but would not spend extra time per patient.” wrote notes on the margin asking if we didn’t understand how busy physicians were, didn’t we know that there was no extra time in the day to do things such as arrange for tests, etc..They indicated that they actually would value the test highly but could not see how they could invest one-half hour or even 10 additional minutes. These responses have encouraged us to believe that the surveyed doctors took the questionnaire seriously and that respondents who volunteered to expend their own time placed a high value in avoiding subjecting children to painful tests.

We also asked respondents how valuable the ability “to obtain an immediate sample at home...at the time of a seizure or adverse event” would be. The positive response was overwhelming. 70% would invest 10 to 30 minutes per patient of their own time “to make advance arrangements for such a test”.

We have no clear explanation why 70% of physicians were willing to expend their own time (question 4) versus 55% for the preceding question. Perhaps the phrase “adverse event” in this last question triggered a stronger response than the routine situation described in question 4. The differing responses between the questions lends credibility to our impression that the child neurologists who responded read each questions carefully and attempted to give an answer that reflected their best clinical judgment.

The ability to obtain an immediate sample in the home is a potential advantage of saliva monitoring.{Bailey, Klein, et al. 1997 936 /id}{Tal, Aviram, et al. 1990 937 /id} This use has not been widely explored but the simplicity of sample collection and the stability of specimens at room temperature suggests it could be practical.{Rosenthal, Hoffer, et al. 1995 81 /id} Especially with epilepsy where seizures can occur infrequently and remembering to adhere to medication schedules can be difficult{Leppik 1998 944 /id} this option could be especially valuable{Stanaway, Lambie, et al. 1985 199 /id}. In the event that a child has a seizure or possible adverse effect, physicians often need an anticonvulsant concentration before suggesting a medication adjustment. With serum this entails a visit to a laboratory or during nights and weekends it means either an emergency department visit or waiting until the laboratory resumes operation. The emergency department visit may not be clinically necessary and incurs further expense. Delaying collection of the sample allows the serum concentration to change so that the measured serum concentration will no longer represent the concentration that existed when the event occurred.

Our data suggest that saliva anticonvulsant monitoring has not been widely adopted. Among American child neurologists, the subjects of this survey, only two respondents had ordered a saliva anticonvulsant level in the last year as part of their routine office practice and less than 7% (37/544) reported even knowing of a laboratory which could perform the test. The child neurologists are the pediatric sub-specialists with the greatest experience in managing epilepsy. Children with epilepsy constitute a large percentage of their patients. If they don’t use an epilepsy related test and don’t even know where to obtain it, in our opinion, it is doubtful if any other group of American physicians caring for children with epilepsy utilizes it.

Our survey suggests that a painless method of measuring anticonvulsant concentrations, especially if the technique would lend itself to obtaining samples at home or school would fill an important and unmet need in the management of children with epilepsy. Using saliva in place of serum might fill that need. Data supporting saliva measurements for phenobarbital, phenytoin and carbamazepine have been available for over 10 years.{Miles, Tennison, et al. 1990 17 /id}{Miles, Tennison, et al. 1991 120 /id} Recent studies have demonstrated close correlations between serum and saliva concentrations for lamotrigine{Trnavska, Krejcova, et al. 1991 123 /id}, levetiracetam{Grim, Ryan, et al. 2003 2471 /id} and topiramate{Miles, Tang, et al. 2003 2474 /id}. It is possible that some clinicians also routinely monitor hepatic, hematopoietic, and other factors which require blood samples. Given the lack of value of such monitoring with the above anticonvulsants, it is difficult to see that this will be a major issue.{Leppik 1998 944 /id}{Leppik, jacobs, et al. 1990 2442 /id}{Wyllie & Wyllie 1991 2441 /id}

The technique for obtaining saliva is easy enough, the child simply spits into a plastic cup.{Gorodischer, Burtin, et al. 1997 57 /id} For infants or children who are unable to cooperate, a simple, disposable pipette can be used to obtain saliva. A drop of citric acid will stimulate saliva production if the child’s mouth is dry without altering the assay.{Gorodischer, Burtin, et al. 1997 57 /id} The major technological problem is obtaining saliva too close in time to oral administration of medication - traces of medication may remain in the mouth and contaminate the sample. A three hour interval is sufficient to avoid this problem.{Dickinson, Hooper, et al. 1985 200 /id}

Our survey data strongly suggest that there is an unmet need for obtaining anticonvulsant concentrations without causing pain - especially if this method could be used at home or school. It is clear that the current leading technology which could meet this need, saliva concentration determinations, has not been adopted. This doesn’t appear to reflect a lack of physician interest. We don’t have data regarding other factors which might inhibit the adoption of this technology. It is possible that there are economic factors which are beyond physician control, such as the loss of revenue from phlebotomy or the cost to the laboratory of switching from one test to another, have favored the status quo. The lack of proprietary methods that would be promoted by patent holders could also be a factor.{Berndt 2001 68 /id}

This questionnaire was limited to a single page. While this is likely to have improved our response rate it certainly limited the number and complexity of the questions that could be asked. In addition, the questions asked respondents to report on their current or future behavior. We have no way of determining whether those who predict they would use a test in the future would actually behave in accordance with their responses. We are encouraged by the distribution of responses that we have avoided simply triggering a socially correct response.{Baumann, Wilson, et al. 1995 2446 /id}

Conclusion:

When queried many American child neurologists are willing to expend their own time to arrange for painless anticonvulsant concentration monitoring for their pediatric patients. An even higher value is placed on obtaining anticonvulsant concentrations without travel to a medical facility if the child should have a seizure or a potentially adverse event. These responses suggest that the circumstances are appropriate to move away from serum-based anticonvulsant concentration monitoring and for the wide spread adoption of saliva or some similar method for monitoring anticonvulsant concentrations for children. This becomes more practical with the increasing numbers of anticonvulsant medications which do not require serum surveillance of hematopoietic, liver or renal function.

Table: Responses to questions 3 and 4 about perceived value of test

*phrase in parenthesis only used in question 4