The impact of the withdrawal of Adderall XR (long-acting mixed amphetamine salts) from the Canadian market on paediatric patients and their families

Abstract

BACKGROUND:

Adderall XR (Shire BioChem Inc, Canada), a medication used to treat attention deficit hyperactivity disorder, was withdrawn from the Canadian market in February 2005 due to concerns of possible cardiotoxicity and cerebral vascular events among a small number of individuals who had taken the medication.

OBJECTIVE:

The primary objective of the present study was to investigate the degree to which the physician’s relationship with the families of the patients to whom the medication was prescribed was affected by the withdrawal of Adderall XR from the Canadian market. The study sought to explore the perceptions of caregivers of patients who took Adderall XR to the drug recall. As a secondary objective, the study also assessed the differences in perception of caregivers toward physicians compared with their perception of other agencies involved with the recall.

METHODS:

Questionnaires were sent to the caregivers of 123 patients who had been taking the drug at the time of the withdrawal.

RESULTS AND CONCLUSIONS:

Of the 53 (43%) completed questionnaires, 89% of respondents indicated that they were concerned when informed of the withdrawal, while 58% indicated that they were frightened. Despite the concerns, only a modest degree of anger was expressed. Thirty per cent of respondents reported anger directed at Health Canada, 24% reported anger directed at the manufacturer, while no caregiver reported anger directed at their physician. Only three families (5.7%) indicated a decrease in confidence in the physician following the event. Fifty-eight per cent indicated a willingness to resume taking Adderall XR, if it was deemed safe by Health Canada. These results offer insight into patient and family perspectives following an unexpected medication recall. While caregivers were generally concerned and often frightened by this event, the present data do not suggest that the parent-physician relationship was greatly affected.

Physicians managing children and adolescents with mental health issues are often called on to prescribe medications, which in themselves, represent significant concerns. Warnings about potentially serious side effects relating specifically to the paediatric population have been placed on antidepressants since May 2004 (1,2). More recently, there have been concerns expressed about the safety of stimulant medications. Little has been published about the perception of patients and their families when faced with these issues.

In February 2005, the proprietary medication Adderall XR (Shire BioChem Inc, Canada) (mixed amphetamine/dextroamphetamine salts) was withdrawn from the Canadian market by Health Canada (3). The concerns were based on 20 international case reports of sudden deaths related to cardiotoxic and cerebral vascular events in patients taking Adderall XR or Adderall immediate-release formula. Fourteen of the 20 cases occurred in the paediatric population. Health Canada later reconsidered its decision, and the medication was returned to the market in November 2005 (4). The present study was undertaken primarily due to concern that this medication withdrawal could adversely affect the clinical relationship between the physician, the patient and their caregivers. In particular, the study was interested in determining how much anxiety these events provoked in families, and whether these events diminished the trust between physicians and the families. To this end, a questionnaire was developed and distributed to caregivers of patients affected by this withdrawal to explore the effect of the drug recall on patients and their families. The study was completed during the period between the recall and return of Adderall XR to the Canadian market. The authors hoped to gain some insight into how these events changed families’ views of the health care system, health care providers and pharmacotherapy for attention deficit hyperactivity disorder (ADHD).

METHODS

Study participants were identified from the practices of consulting general paediatricians in a secondary centre (Chatham-Kent Health Alliance, Chatham, Ontario). All three paediatricians in the centre participated in the study. Each paediatrician practiced independently and maintained his/her own roster of patients. None of the paediatricians provided any primary care, only consultative services were provided. The vast majority of patients were referred from local family physicians.

The individual paediatricians assessed referred patients, and a diagnosis was assigned using criteria from the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (5). The methods of assessment varied from paediatrician to paediatrician, but generally included a clinical interview, questionnaires filled out by caregivers and school personnel, and standardized ADHD symptom checklists filled out by parents and teachers. Many of the children also underwent an informal screening of skills (motor, language, memory, visual or perceptual skills) by the paediatrician, which was approximately 1 h in length. Children started on medication were generally followed-up every four to six weeks until stabilized on a dose of medication. Once stabilized, cases were generally reviewed every three to six months. Children experiencing difficulties were seen more frequently. A minority of patients were referred for assessment by a paediatric psychiatrist. These were generally completed by video link, and usually consisted of a single consultation. A small number of the most severe cases were referred to larger centres for residential care. After discharge from the residential centres, the local paediatricians would resume care of the patient.

When the announcement from Health Canada was made indicating that Adderall XR had been withdrawn, office records were reviewed and all affected patients were identified. These patients were phoned directly by either the paediatrician or by the office staff. All patients were called within 36 h of the announcement of the medication withdrawal. Patients were instructed to stop taking the medication. Follow-up was arranged or alternative treatment was offered on a case-by-case basis. The opportunity for a research project was recognized immediately, and the list of affected patients was retained for eventual use after the project was approved.

A questionnaire was developed, requesting patient demographics and medication history. Caregiver perceptions were assessed through the use of a series of questions, with responses given in the format of Likert scales with five degrees of agreement.

Questionnaires were mailed to the caregivers of all patients (n=123) who had been taking Adderall XR at the time of the medication withdrawal. Respondents were assured that their clinicians would not be aware of their responses. Questionnaires were coded such that the authors directly involved in clinical care were unaware who had responded or how any individual had responded. Families who did not respond were sent a reminder approximately two weeks after the initial contact. Questionnaires were mailed in early September 2005 and final responses were received in late October 2005.

The nonclinician investigator reviewed the charts of those agreeing to participate in the study, and information was collected on patient demographics, diagnosis, comorbidities and medication history.

Whether more anger was directed at the physician or the pharmaceutical company was assessed using the Wilcoxon’s signed rank test. Similarly, anger directed at the physician versus Health Canada was also assessed. Because the questionnaires were not validated, a nonparametric test comparing responses with paired questions was thought to be most appropriate. During post hoc analysis, it was noted that families whose children were not able to find a suitable replacement for Adderall XR seemed more likely to be receptive to resuming the medication if it became available than families who did find an adequate replacement. Patients doing at least as well as previously were compared with those doing less well with regard to their willingness to resume Adderall XR. The significance of this trend was assessed using the McNemar χ² test. A biostatistician reviewed the results.

The existing literature regarding patient perspectives on medication recalls was reviewed through a PUBMED search (major MeSH term ‘Drug and Narcotic Control’ or ‘United States Food and Drug Administration’ combined with title/abstract terms ‘withdrawal’ or ‘recall’; major MeSH term ‘Drug and Narcotic Control’ combined with major MeSH term ‘Physician-patient relations’), followed by a manual review of the relevant citations.

The research ethics board of the University of Western Ontario (London, Ontario) approved the project. All participants included gave informed consent for the chart review.

RESULTS

Out of 123 questionnaires, 55 were returned. Two questionnaires were returned blank, leaving 53 (43%) for analysis. Patient demographics along with details regarding time since diagnosis and medication history are presented in Table 1. A more detailed medication history is described in Table 2.

TABLE 1.

Patient demographics (n=53)

Characteristic	Results
Age at Adderall XR* withdrawal, years (mean [range])	11.64 (5–18)
Male sex, %	86.8
Time between ADHD diagnosis and Adderall XR withdrawal, weeks (mean [range])	46.1 (2–168)
Duration of Adderall XR use, weeks (mean [range])	20.4 (1–48)
Number of medications tried before Adderall XR for behaviour, mean (range)	2.95 (0–8)
Number of medications tried after Adderall XR withdrawal for behaviour, mean (range)	1.43 (0–4)

*Shire BioChem Inc, Canada. ADHD Attention deficit hyperactivity disorder

TABLE 2.

Parental reporting of medications used (n=53)

Medication	Number of patients prescribed the following medications for behaviour/school-related issues before trying Adderall XR*	Number of patients prescribed the following medications as an alternative to Adderall XR since its removal from the market
Methylphenidate immediate-release	33	9
Methylphenidate intermediate-release (Ritalin SR†)	17	3
Methylphenidate sustained-release (Concerta‡)	28	27
Dextroamphetamine	21	12
Dextroamphetamineintermediate-release (Dexedrine spansule§)	19	13
Atomoxetine (Strattera¶)	0	10
Clonidine	13	Not applicable
Risperidone	15	Not applicable
Olanzapine	1	Not applicable
Buproprion	2	Not applicable
Fluoxetine	1	Not applicable
Sertraline	1	Not applicable
Paroxetine	1	Not applicable
Fluvoxamine	1	Not applicable
Citalopram	2	Not applicable
Homeopathic medicines	3	2

*Shire BioChem Inc, Canada;

^†Novartis Pharmaceuticals, Canada;

^‡Janssen-Ortho Inc, Canada;

^§GlaxoSmithKline Inc, Canada;

^¶Eli Lilly Canada Inc

Table 3 illustrates caregiver responses to Adderall XR withdrawal. The vast majority of patients indicated that they were concerned when informed of the withdrawal, with a substantial number agreeing that they were actually frightened. Despite the obvious concerns, no respondent indicated that they were angry with the prescribing physician, and only 5.7% admitted to feeling less confident in accepting the advice of the clinician after the medication withdrawal. Any anger that existed was more likely to be directed at the pharmaceutical company manufacturing Adderall XR than at the physician (Z=–4.83, P<0.001, Wilcoxon’s signed rank test). Similarly, anger was more likely to be directed at Health Canada than at the physician (Z=–5.41, P<0.001, Wilcoxon’s signed rank test). Despite the concern at the time of the withdrawal, many families (58.4%) indicated a willingness to resume Adderall XR use if it was deemed safe by Health Canada. Families whose child was unable to find a satisfactory alternative to Adderall XR were significantly more likely to express willingness to resume the medication compared with families who had found a satisfactory alternative (χ²=14.2, P<0.01, McNemar χ² test). All but one of the 20 families who were unable to find another medication as effective as Adderall XR indicated a willingness to resume the medication if it became available again.

TABLE 3.

Caregiver responses to the Adderall XR (Shire BioChem Inc, Canada) withdrawal from the Canadian market (n=53)

Questions	%
While it was available, Adderall XR worked well for my child. Consider both effects and side effects
Strongly agree	43.4
Agree	30.2
Neither agree or disagree	9.4
Disagree	13.2
Strongly disagree	3.8
If Health Canada felt that it was safe to market Adderall XR in this country once again, I would feel comfortable having my child try it again if the paediatrician recommended it
Strongly agree	34.0
Agree	24.5
Neither agree or disagree	9.4
Disagree	13.2
Strongly disagree	17.0
Overall, how do you feel that your child is doing now compared to when he/she was taking Adderall XR?
Much better	13.2
Somewhat better	11.3
About the same	35.8
Somewhat worse	30.2
Much worse	7.5
When I heard that Adderall XR had been withdrawn, I was concerned
Strongly agree	64.1
Agree	24.5
Neither agree or disagree	5.7
Disagree	5.7
Strongly disagree	0.0
When I heard that Adderall XR had been withdrawn, I was frightened
Strongly agree	30.2
Agree	28.3
Neither agree or disagree	11.3
Disagree	26.4
Strongly disagree	3.8
When I heard that Adderall XR had been withdrawn, I was angry at Health Canada
Strongly agree	7.5
Agree	22.6
Neither agree or disagree	26.4
Disagree	30.2
Strongly disagree	13.2
When I heard that Adderall XR had been withdrawn, I was angry at the drug company that marketed Adderall XR
Strongly agree	11.3
Agree	13.2
Neither agree or disagree	37.7
Disagree	28.3
Strongly disagree	9.4
When I heard that Adderall XR had been withdrawn, I was angry at the physician who prescribed the medication
Strongly agree	0.0
Agree	0.0
Neither agree or disagree	13.2
Disagree	39.6
Strongly disagree	47.2
I feel that I have been given adequate information as to why Adderall XR was withdrawn from the market
Strongly agree	32.1
Agree	32.1
Neither agree or disagree	7.5
Disagree	15.1
Strongly disagree	11.3
I am just as confident now in accepting advice regarding medication from my paediatrician as I was before Adderall XR was withdrawn from the market
Strongly agree	66.0
Agree	24.5
Neither agree or disagree	0.0
Disagree	1.9
Strongly disagree	3.8

DISCUSSION

Several medications have been withdrawn from the Canadian or North American market in recent years. Notable among these have been the withdrawal of certain COX-2 inhibitors (rofecoxib) (6), certain gastrointestinal motility agents (cisapride) (7), as well as the withdrawal from the Canadian market of the stimulant medication Adderall XR. While some information is available regarding the issues leading to these withdrawals, little has been published on the effect that these events have on those patients directly involved. Watkins (8) reported patient perspectives on warnings that unopposed estrogen hormone replacement therapy was found to be associated with endometrial cancer. Patient comments were collected informally from a variety of sources. Diverse findings were reported, ranging from a loss of faith in physicians and medications to anger that removal of an effective drug from the market was being considered.

The present study was initiated due to concerns that the withdrawal of a medication which had been prescribed to a patient might adversely affect the clinical relationship between the prescribing physician, the patient and their families. Specifically, it was unclear how much anxiety these events provoked in the families of these patients, and whether this would diminish the trust between physician and family. Reviewing the literature, no other papers were identified that addressed this question directly. While the literature review identified many articles discussing how future drug recalls might better be avoided, only two articles were identified that addressed patients’ perspectives of recalls that had already occurred. Hawker et al (9) surveyed patients’ perspectives on the recall of rofecoxib, but this study did not address the effect on the physician-patient relationship. Rognstad and Straand (10) identified significant fear among patients following the withdrawal of rofecoxib, and showed that patients wanted and expected to be contacted by their physician following such an event.

It was somewhat surprising that the vast majority of respondents did not indicate that the clinical relationship with the prescribing physician was affected by these events. We note that many of these patients were quite treatment-resistant. Most had been tried on a number of different medications before starting Adderall XR (mean 2.95, range zero to eight). As such, they had met with the prescribing physician on a substantial number of occasions before the withdrawal. Most patients had been seen frequently (mean 12.9 visits) over a fairly prolonged period of time (mean time since diagnosis 47 months). It is possible that this relationship served to insulate the physician from anger and distrust on the part of the families involved. Supporting this notion, Concannon and Tang (11) reported a positive correlation between frequency of physician visits and patient satisfaction.

There are several limitations to the present study. First, all patients were drawn from a single geographical area in southwestern Ontario. It is unclear whether these results can be generalized, particularly to more urban populations. Aside from geographical proximity, we have no further data regarding the characteristics of the caregivers responding to the survey. In a larger study, this would be important for characterizing more fully the factors affecting responses to such medication withdrawals. The degree of parental satisfaction with ADHD management has been shown to be influenced by caregiver characteristics such as age, ethnic background and socioeconomic class (12). It is likely that similar traits will influence responses to medication withdrawals. Similarly, we had hoped to describe the study population more completely in terms of comorbidities. Unfortunately, in reviewing the charts retrospectively, it was found that many comorbidities were incompletely described or not formally identified using Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (5) criteria. For this reason, these data were not included in the present paper.

Second, the questionnaire was not validated. Although care was taken to ask clear and direct questions, it is entirely possible that some of those individuals reporting fear or a loss of confidence were, in fact, relaying some other type of discomfort.

Third, with a response rate of 43%, some degree of non-responder bias is inevitable. Given the design of the study, it was not possible to look more closely at the nonrespond-ing group without breaking confidentiality. Nonresponders did not give consent for a chart review by the nonclinician researcher, and the clinicians were unaware who had responded. We would make several further comments regarding this issue. Most surveys involving ADHD patients and their families take place in the clinic setting. Interestingly, we were only able to identify a single publication reporting the results of an anonymous mail-in survey directed specifically at families affected by ADHD. Landgraf et al (13) reported a response rate of 31%, which is substantially lower than the response rate of our study. We did identify one other study (11) using a mail-in survey that examined parental perspectives on ADHD management, but these data were a subset collected during a much larger survey. We also noted that only 12 of 123 patients were lost to follow-up following the medication withdrawal. Considering the high influx and efflux of ADHD patients in these practices, even in the absence of a drug recall, we do not feel that this is an excessive number. This suggests to us that the degree of dissatisfaction with the physicians following the withdrawal was modest.

Finally, because the sample size was relatively small, it was not possible to identify other factors that correlated with trust in the physician. Only three families reported a loss of trust in the physician, making correlations between trust and other factors impossible.

Opportunities exist for further investigation in this area. Studies could include exploring patient perceptions of drug recall in larger urban centres and in different practice settings. Larger studies would be required to identify what patient and/or caregiver characteristics are correlated with adverse effects on the relationship with the physician under unexpected circumstances; such as a medication withdrawal or medication warnings. Follow-up studies could also explore patient and family perspectives on the return of Adderall XR to the market. With recent warnings that have now been applied to all stimulant medications (14,15), caregiver perspectives on ADHD pharmacotherapy could be explored in general.

CONCLUSIONS

The present study suggests that the withdrawal of Adderall XR from the Canadian market caused considerable concern for the families involved. There was no evidence that the medication withdrawal unduly affected the physician-parent relationship in the practice setting described in the present study, although further work will be required to address issues of nonresponder bias.