Abstract
Aims
To develop a systematic generic method of enabling patients to report symptoms which they believe to be due to a particular prescribed drug.
Methods
A piloted body system-based questionnaire was distributed to patients registered with 79 medical practices in Grampian prescribed one of nine recently marketed ‘black triangle’ drugs. These comprised four antidepressants, three antiepileptics and two analgesics. This requested respondents to identify any symptoms experienced over the previous year which they thought could be due to the ‘black triangle’ drug they had used. A sample of medical records was examined to compare symptoms recorded with those reported by patients. A classification system was developed for the study to enable the assessment of symptoms reported for their potential relationship to patients' drug therapy. All symptoms reported were classified, taking into account information provided by patients on their concomitant drugs and diseases. A specialist pharmacist independently re-classified a sample of the symptoms to validate the process.
Results
A 36.3% response rate was obtained (837/2307) with 742 respondents (88.6%) reporting at least one symptom. The median per patient was 6.0 (range 0–71), with almost half (406, 48.5%) reporting fewer than five symptoms. Most symptoms (71.0%) were classified as being probably or possibly related to the drugs studied. Agreement between researcher and specialist on the classification of 75.3% of 716 symptoms was obtained (Kappa = 0.563). Responses from patients prescribed antidepressant drugs were more likely to include symptoms potentially caused by these drugs (74.5% of all symptoms reported) than those from patients prescribed analgesics (67.4%) or antiepileptics (65.1%, χ2 = 23.858, d.f. = 2, P < 0.001). Patients reporting large numbers of symptoms were more likely to report some which were classed as unlikely to be an ADR or unattributable (χ2 = 80.587, d.f. = 3, P < 0.001). Of the 742 reporting symptoms in questionnaires, 402 (54.2%) claimed to have reported some or all of these to their doctor. Only 162 (22.6%) of 716 patient-reported symptoms were documented in the primary care medical records of 103 patients prescribed tramadol or venlafaxine.
Conclusions
Respondents were clearly willing to report symptoms, the majority of which were classed as possibly/probably related to the drugs studied. The results suggest that patients do not report all symptoms they suspect to be ADRs to their GP and that GPs do not record all symptoms which may be reported to them. The method could help to identify problems which patients perceive as being related to their drug therapy and contribute to increased ADR reporting.
Keywords: ADR classification, adverse drug reactions, patient self-reporting
Introduction
Patients experience undesirable effects from drugs despite taking them as directed [1]. Clinical trials frequently fail to identify all ADRs [2], in fact 51% of serious ADRs of marketed drugs were not detected prior to approval [3]. These limitations have created a need for comprehensive, efficient and cost-effective systems for postmarketing drug surveillance that confirm or extend the data from premarketing studies and determine the range of ADRs in the general population [4, 5]. The contribution of ADRs to hospital admission has been highlighted [6, 7]. In another study community-acquired ADRs represented 42% of the total 541 reported ADRs and 87% of the cost of treating them [8].
Spontaneous reporting systems such as the UK's Yellow Card scheme serve as an early warning system for serious and unexpected ADRs. These depend on voluntary reporting of suspected reactions by health professionals. In the UK this is either directly to the Committee on the Safety of Medicines (CSM) or indirectly through pharmaceutical companies [9–12]. Under-reporting is a major problem [13, 14] estimated at 90% [15]. Other pharmacovigilance systems in use in the UK are prescription-event monitoring [16], which also suffers from low GP response rates [17], the General Practice Research Database and the record linkage system in Tayside [18].
The CSM requests that all suspected reactions are reported to newly marketed medicines which are identified in standard texts with a black inverted triangle. Such drugs are known widely throughout the UK as ‘black triangle’ drugs. Recently, the UK Yellow Card scheme was extended to enable pharmacists to report directly to the CSM [19]. Reports submitted by hospital pharmacists in the first year included a lower proportion of reactions related to ‘black triangle’ drugs compared to hospital doctors [20]. Few community pharmacists in one survey knew the reporting criteria for these drugs [21].
Few studies on ADR reporting have involved patients directly, yet both the Yellow Card scheme and prescription event monitoring rely in part on the patient spontaneously reporting symptoms to health care professionals. There is little published literature regarding the development of formal patient-initiated surveillance approaches [22]. The national ADR registers in the USA and Germany accept reports directly from patients [23]. Encouragement to report symptoms to GPs by means of a leaflet issued by community pharmacists found an increase in the number of symptoms reported [24], but did not increase Yellow Card reports. A similar method of encouragement, but for reporting to a research centre, developed in the USA found that common, well-known ADRs were reported [25–27]. The researchers concluded that patient self-monitoring could be a promising complement to existing doctor-based reporting systems as well as a possible early alerting mechanism for detecting ADRs to new drugs. An Australian study using symptom lists distributed by community pharmacists found that patient self-reporting identified established ADRs, suggesting that the system might generate early warning of symptomatic reactions to new drugs [28]. This is supported by work comparing reports from patients and professionals [29].
Patients' involvement in treatment decisions is increasing [30], yet their experiences of using medicines are not routinely sought. For this reason, the present study involved the development of a generic method to enable patients to report symptoms which they believed to be due to a particular prescribed drug, with a focus on newly marketed products.
Methods
Local ethics committee approval was obtained for the study.
Questionnaire development
Questionnaires were developed, using previously published work as a basis [25, 31], which sought information from patients on current drug therapy, disease states and symptoms experienced which could be due to medicines, using both open and closed alternatives. These questionnaires were piloted by sending to a small number of patients (14) registered with one medical practice with a request to complete all questions. The 11 patients who completed the questionnaires were subsequently interviewed to determine their ability to complete the different types of questions. Further development of the questionnaire was based on the views of these patients.
The questionnaire was further piloted in a sample of patients receiving well-established drugs. Minor changes were made as a result of this pilot study.
The final questionnaire (see supplementary material on the Internet; please refer to end of references) consisted of three sections:
open and closed questions seeking demographic details and information on concurrent therapy and disease states;
an extensive series of symptoms divided into body systems or regions, with tick boxes and spaces for addition of other symptoms; an open question to elicit the most bothersome symptoms of those ticked; closed questions relating to symptom severity and reporting of symptoms to the GP;
closed questions on reasons for stopping the index drug and space to record any symptoms appearing after stopping.
Questionnaire distribution
Pilot study: Questionnaires were posted to patients from three medical practices who were identified from practice computer databases as being prescribed one of five drugs: carbamazepine, sodium valproate, doxepin, trazodone and coproxamol. One reminder was sent to nonresponders. Results from this pilot have been reported [32].
Main study: All 97 general practices in the Grampian region of Scotland were approached by letter to participate in the study, accompanied by a prepaid envelope for return if they did not wish their patients to be involved. The remaining 79 practices were sent a reminder letter and sample questionnaire 2 weeks prior to the distribution of questionnaires to patients. Patients registered with the participating practices who were prescribed one of nine ‘black triangle’ drugs during the period January to March 1997 (venlafaxine, nefazodone, citalopram, moclobemide, gabapentin, topiramate, lamotrigine, tramadol, fentanyl patch) were identified via prescription details obtained from the Pharmacy Practice Division (PPD). A printout of identifier numbers obtained from PPD enabled identification of prescriptions for the drugs listed. Searching through original prescription forms provided patients' names and addresses.
Questionnaires were sent to all patients identified by this process excluding those under 16 years old and those resident in nursing homes. In view of the nature of the drugs prescribed and the possibility of anxiety being generated by receipt of a questionnaire, no reminders were issued.
The number of Yellow Card reports sent to the CSM by participating practices concerning these drugs over the study period were supplied by the Medicines Control Agency for comparison.
Classification of symptoms reported
As limited data only were available from patients' questionnaires, a new method of classifying each symptom was developed which accounted for the causal relationship between symptoms reported and study drugs, concomitant drugs, or disease states. This method allowed symptoms to be grouped into four categories: probable ADR, possible ADR, unlikely to be an ADR, symptoms previously not attributed to study drugs and unattributable to other drugs or disease states. Criteria for each category are shown in Table 1. All symptoms reported were classified, taking into account the information provided by patients on their concomitant drugs and diseases. The literature on ADRs was used in this process, to identify whether each symptom was previously described as an ADR to any of the drugs and standard texts were used in attributing symptoms to disease states. The purpose of this classification system was to enable the usefulness of the symptom reports to be evaluated.
Table 1.
Classification criteria used for symptoms reported by patients.
| 1. Probable – symptom probably caused by the index drug known or previously reported reaction to index drug and could not reasonably be explained by the effects of concomitant drug(s) and could not reasonably be explained by the known characteristics of the patient's clinical condition |
| 2. Possible – symptom possibly caused by the index drug known or previously reported reaction to index drug and could reasonably be explained by the effects of concomitant drug(s) and/or by the known characteristics of the patient's clinical condition |
| 3. Unlikely – symptom unlikely to be caused by index drug: not known or not previously reported reaction to index drug and could reasonably be explained by the effects of concomitant drug(s) and/or by the known characteristics of the patient's clinical condition |
| 4. Unattributable: not known or not previously reported reaction to index drug and could not reasonably be explained by the effects of concomitant drug(s) and could not reasonably be explained by the known characteristics of the patient's clinical condition |
Validation
Medical records of all respondents in the pilot study who reported a symptom were examined to compare information with that provided by patients. The same process was used in a sample of patients reporting symptoms to venlafaxine or tramadol. The symptom classifications made by one of the researchers (N.J.) for all symptoms reported by these latter patients were compared to those made independently by a pharmacist with expertise in pharmacovigilance (A.L.), using the same classification.
Analysis of data
Data were analysed using SPSS for Windows version 9.0 and Minitab version 12. Numbers of symptoms and concomitant drugs reported were further subgrouped to enable relationships between these and other variables to be investigated, using chi-squared tests. The 95% confidence interval or P value at 0.05 was chosen to accept or reject the null hypotheses. Comparison of classifications was analysed using the Kappa statistic [33].
Results
Response rates and demographic data
A total of 2307 postal questionnaires were distributed in the main study, from which 837 valid responses were obtained (36.3% response), plus 87 responses considered invalid, due to lack of completion or failure to take the prescribed index drug. Response rate differed between drugs, being greatest for antiepileptic drugs, although the actual numbers were small (Table 2). Most questionnaires were issued to and returned from patients prescribed tramadol and venlafaxine, the most frequently prescribed drugs.
Table 2.
Response rates and symptoms reported for each drug.
| Drug | Number of questionnaires sent | Number of valid respondents (%) | Number of different symptoms reported | Total number of symptoms reported (median, range) |
|---|---|---|---|---|
| Fentanyl patch | 64 | 8 (12.5) | 47 | 94 (12.5, 2–24) |
| Topiramate | 23 | 13 (56.5) | 41 | 59 (3.0, 0–12) |
| Moclobemide | 48 | 16 (33.3) | 52 | 103 (6.5, 0–16) |
| Lamotrigine | 40 | 18 (45.0) | 46 | 117 (5.0, 0–17) |
| Citalopram | 132 | 43 (32.6) | 78 | 358 (5.0, 0–37) |
| Nefazodone | 204 | 64 (31.4) | 85 | 677 (8.5, 0–39) |
| Gabapentin | 115 | 68 (59.1) | 77 | 575 (5.5, 0–37) |
| Venlafaxine | 633 | 263 (41.5) | 97 | 2700 (7.0, 0–71) |
| Tramadol | 1048 | 344 (32.8) | 92 | 2333 (4.5, 0–51) |
The majority of respondents were female (66.2%) and 0.6% did not specify. The average age was 50.5 years (s.d. 17.2 years), with the majority of respondents falling into the 40–59 (38.0%) and 20–39 (28.2%) age groups and there were no differences between the genders with respect to age. Approximately half the respondents (420, 50.2%) claimed to be taking between 1 and 3 prescribed medicines in addition to the index drug, 271 (32.3%) listed more than this number, with the remainder either stating none or not listing them.
The medical records of 310 respondents in the pilot and main studies were examined. Differences were found in 218 (70.3%) cases between the medicines stated by patients in the questionnaires as being taken and the details found in their records. Data on disease states provided in the questionnaires were also frequently incomplete, with only 26 (8.3%) cases having records and questionnaire data in full agreement.
Number of symptoms reported
Of the total 837 respondents, 742 (88.6%) reported at least one symptom. The total number of symptoms reported was 7016 and the median per patient was 6.0 (range 0–71). Patients prescribed tramadol and venlafaxine reported a wider range of symptoms compared with those prescribed other drugs, while the greatest median number of symptoms reported came from those prescribed fentanyl patch (Table 2). Almost half the respondents (406, 48.5%) reported fewer than five symptoms. There were no significant relationships between the number of symptoms reported and the number of concomitant medicines reported or duration of therapy.
Classification of symptoms
Fifty-five percent (3848) of the total number of perceived symptoms reported were classified by N.J. as being possibly caused by the drugs studied, with a further 1134 (16.2%) classed as probably caused by these drugs. Seventeen percent (1226) were classed as unlikely to be an ADR and 11.6% (808) were unattributable to either drugs or disease states on the basis of the details provided by respondents.
A total of 716 symptoms reported by a sample of 103 patients taking venlafaxine or tramadol were classified independently by A.L. There was agreement on the classification of 75.3% of these symptoms, resulting in a Kappa value of 0.563 and indicating a moderate level of agreement. Further analysis was carried out using this classification system.
For all the drugs studied, symptoms which the researcher (N.J.) classified as potentially related to the index drug (probable/possible) were reported more frequently than symptoms classed as not likely to be related to the index drug (unlikely/unattributable) (χ2 = 148.632, d.f. = 8, P < 0.001) (Table 3). Patients reporting symptoms to antidepressant drugs were more likely to report symptoms considered related to these drugs (74.5% of all symptoms reported) than patients taking analgesics (67.4%) or antiepileptics (65.1%, χ2 = 23.858, d.f. = 2, P < 0.001). The greater the number of symptoms reported, the greater the likelihood of some of these being classed as unlikely to be an ADR or unattributable (χ2 = 80.587, d.f. = 3, P < 0.001) (Table 4).
Table 3.
Patient attribution accuracy of symptoms as side-effects of drugs.
| Drug | Frequency of yes criteria* (%) | Frequency of no criteria** (%) | Total (%) |
|---|---|---|---|
| Venlafaxine | 2067 (76.6%) | 633 (23.4%) | 2700 (100%) |
| Nefazodone | 408 (60.3%) | 269 (39.7%) | 677 (100%) |
| Citalopram | 303 (84.6%) | 55 (15.4%) | 358 (100%) |
| Moclobemide | 80 (77.7%) | 23 (22.3%) | 103 (100%) |
| Gabapentin | 387 (67.3%) | 188 (32.7%) | 575 (100%) |
| Lamotrigine | 70 (59.8%) | 47 (40.2%) | 117 (100%) |
| Topiramate | 32 (54.2%) | 27 (45.8%) | 59 (100%) |
| Tramadol | 1567 (67.2%) | 766 (32.8%) | 2333 (100%) |
| Fentanyl patch | 68 (72.3%) | 26 (27.7%) | 94 (100%) |
| Total | 4982 (71.0%) | 2034 (29.0%) | 7016 (100%) |
yes criteria = reported symptoms potentially related to index drugs (probable/possible)
no criteria = reported symptoms not likely to be related to index drugs (unlikely/unattributable).
Table 4.
Unlikely ADRs and unattributable symptoms in relation to number of reported symptoms.
| Number of symptoms reported | Total number of patients | Number (%) of patients reporting ≥1 unlikely ADR | Number (%) of patients reporting≥1 unattributable symptom |
|---|---|---|---|
| 1–5 | 313 | 79 (25.2%) | 107 (34.2%) |
| 6–10 | 171 | 78 (45.6%) | 119 (69.6%) |
| 11–15 | 110 | 79 (71.8%) | 94 (85.5%) |
| > 15 | 148 | 131 (88.5%) | 137 (92.6%) |
Symptom severity and reporting
A total of 585 respondents provided a rating for the severity of their most bothersome symptoms, of whom 231 (39.5%) viewed them as moderately severe, with a further 251 (42.9%) giving a rating of severe or very severe. The remaining 103 (17.6%) gave ratings of mildly or minimally bothersome. Although no significant association was found between the total number of concomitant medicines and symptom severity (χ2 = 12.776, d.f. = 8, P = 0.120), there was a trend towards those taking more medicines rating their symptoms as moderate or severe (Figure 1). Increasing severity ratings were also associated with increasing numbers of symptoms reported. (χ2 = 75.765, d.f. = 8, P < 0.001)
Figure 1.
Severity ratings of most bothersome symptoms in relation to number of symptoms reported.
There were 216 patients who indicated they had reported all their symptoms to their doctor and a further 186 who reported some symptoms. The majority of these who also gave ratings for symptom severity (337, 83.8%) rated their most bothersome symptom as moderate or severe (χ2 = 43.926, d.f. = 2, P < 0.001). Examination of the medical records of 103 patients prescribed venlafaxine or tramadol found documented evidence of only 162 (22.6%) of the 716 symptoms the patients reported in the questionnaires. Of these, 27 were documented as possible side-effects of the study drugs and 7 as possible side-effects of concomitant therapy. Although the 607 patients receiving these two drugs reported a total of 5033 symptoms, only 23 reports were received by the CSM of adverse events relating to venlafaxine or tramadol over the same study period from the 79 participating practices.
Discussion
The questionnaire was designed to enable patients to report a wide range of symptoms to any class of drug and was distributed directly to patients, with no reminder. The overall response rate was 40.1%, with a usable response of 36.3%. This was lower than the 44% found in the pilot study, in which a reminder was issued. Distribution of questionnaires to enable ADR reporting by pharmacists in Australia found a similar (39%) response rate [34]. Pharmacists recruiting patients to a prospective study self-reporting in relation to NSAIDs in Canada obtained the agreement of 51% of their patients of whom 83% (42% of those approached) subsequently provided data [35].
Respondents were clearly willing to report symptoms which they perceived could be due to the drugs being taken using this questionnaire. It has already been suggested that self-reports can be a useful added source of information on minor ADRs which are important to patients [34]. Patients were also found to report perceived ADRs more rapidly than health professionals [29]. Experiencing a perceived ADR may be a contributory factor to failing to take medicines and detection is thus important.
Symptom checklists could increase reporting rates through suggestion [36], although one study found this was not the case [37]. Many checklists which have been developed for individual drugs show that patients consistently identify well-recognized symptoms [38–41]. However open questionnaires, while not suggesting possible ADRs, identify fewer symptoms [42, 43]. Spontaneous reporting by patients appears to be more reliable for detecting unexpected ADRs than systematic enquiry or checklists [44], but may also result in a lower reporting rate [45].
Shorter, drug-specific checklists result in similarly high proportions of respondents reporting symptoms to the 88.6% found in the present study [46, 47], whereas many fewer respondents identify symptoms in techniques which use open questions [27, 35]. The questionnaire checklist may also result in a large number of symptoms which are not attributable to the drugs being studied. Only 11.5% of symptoms in this study could not be attributed to drug therapy or diseases, similar to the proportion found in the pilot study [32]. Significantly more symptoms reported were classed as potentially related to the index drugs, although the majority of these could also have been due to concomitant therapy or to reported disease states (possible ADR). A similar proportion of symptoms were classed as probable or potential ADRs in the pilot study [32] and is in agreement with that found elsewhere [26]. These results suggest that patients are most likely to report genuine symptoms which have a high probability of being drug-related.
Many unattributable symptoms and those classed as unlikely to be ADRs were reported by patients who identified large numbers of symptoms. Similarly, patients who reported many symptoms were more likely to rate their most bothersome symptoms as severe. Patients' tendency to report symptoms which were severe and bothersome to them has also been found elsewhere [27, 48]. There was a trend towards higher numbers of concomitant drugs being taken by patients who claimed to have severe symptoms. The use of three of more drugs was a significant factor in one study of patients with severe adverse reactions admitted to hospital [49]. Many claiming severe symptoms also claimed to have reported some or all of their symptoms to their doctor. Only 22% of the symptoms reported by a selection of patients were found documented in primary care medical records, which was again in agreement with the pilot [32]. It is possible therefore, that patients' failure to report symptoms could be a contributory factor to low levels of spontaneous reporting, since all methods rely to an extent on the patient informing their doctor. This is reflected in the small number of formal reports submitted on the drugs studied from the participating practices.
One of the limitations of patient self-reporting is that data are obtained from patient perceptions and recollections. While patients may not be regarded as able to discriminate effectively between symptoms which are attributable to individual drugs or diseases, there are similar problems among health professionals [10, 50]. Despite the creation of various criteria to assess the causality of ADRs [51–53], such methods still only categorize ADRs into levels of probability. The need to develop a new classification system to evaluate the patients' reports, based on different criteria, arose because of the lack of temporal data available to enable any of these standard criteria to be used. Also the incompleteness and potential inaccuracy of data provided by patients on concomitant therapy and diseases would have contributed to difficulties in attributing symptoms. However despite these limitations, the results suggest that most of the symptoms reported by patients were potentially drug-related.
While there are clearly difficulties in using a checklist, if used in a screening process by a health professional, such a method could enable the identification of problems which patients perceive as being related to their drug therapy. One potential benefit of this is increased reporting to authorities of the symptoms reported. The results suggest that it would also be of help in identifying common symptoms which patients find bothersome. Another potential use of the checklist is in the process of reviewing individual patients' medication as part of pharmaceutical care [54]. A generic symptom checklist would be useful in identifying drug-related problems and also in identifying untreated symptoms [55]. Furthermore, a body system-based comprehensive symptom checklist has recently been advocated for use in clinical trials [56]. The checklist reported here has been found to be acceptable to patients and resulted in many symptoms being reported. The study suggests that patient reports of potential ADRs may be of value for postmarketing surveillance since the symptoms reported were seldom unattributable to drug therapy. However further work is needed to establish an effective mechanism by which this could be used to alert health professionals to these symptoms.
Acknowledgments
We would like to thank all practices for allowing us to send questionnaires to their patients and all the patients who returned them. We are also grateful to the Pharmacy Practice Division (now Information and Statistics Division) for assistance in identifying patients, to the Medicines Control Agency for supplying data and to the Ministry of University Affairs, Thailand, for financial support for N.J.
Supplementary Material on the Internet
The questionnaire used in this study is available from http://www.blackwell-science.com/products/journals/suppmat/BCP/BCP1547/BCP1547sm.htm
References
- 1.Einarson TR. Drug-related hospital admissions. Ann Pharmacother. 1993;27:832–840. doi: 10.1177/106002809302700702. [DOI] [PubMed] [Google Scholar]
- 2.Ray WA, Griffin MR, Avorn J. Evaluating drugs after their approval for clinical use. N Engl J Med. 1993;329:2029–2032. doi: 10.1056/NEJM199312303292710. [DOI] [PubMed] [Google Scholar]
- 3.Moore TJ, Psaty BM, Furberg CD. Time to act on drug safety. JAMA. 1998;279:1571–1573. doi: 10.1001/jama.279.19.1571. [DOI] [PubMed] [Google Scholar]
- 4.Generali JA, Danish MA, Rosenbaum SE. Knowledge of and attitudes about adverse drug reaction reporting among Rhode Island pharmacists. Ann Pharmacother. 1995;29:365–369. doi: 10.1177/106002809502900404. [DOI] [PubMed] [Google Scholar]
- 5.Timm EG, Welage LS, Walawander CA, et al. Adverse drug reaction reporting in a multicenter surveillance study. Ann Pharmacother. 1995;29:240–245. doi: 10.1177/106002809502900302. [DOI] [PubMed] [Google Scholar]
- 6.Cunningham G, Dodd TR, Grant DJ, McMurdo MET, Richards RME. Drug-related problems in elderly patients admitted to Tayside hospitals, methods for prevention and subsequent reassessment. Age Ageing. 1997;26:375–382. doi: 10.1093/ageing/26.5.375. [DOI] [PubMed] [Google Scholar]
- 7.Pongwecharak J. Monitoring Medication Problems in the Elderly and Strategies for Their Reduction. The Robert Gordon University; 1998. PhD Thesis. [Google Scholar]
- 8.Clyne KE, German MR. Adverse drug reaction reporting. Focus on cost and prevention. Pharm and Ther. 1992;17:1145–1156. [Google Scholar]
- 9.Mann RD. Drug safety alerts – a review of ‘current problems’. Pharmacoepidemiol Drug Safety. 1992;1:269–279. [Google Scholar]
- 10.Bateman DN, Sanders GL, Rawlins MD. Attitudes to adverse drug reaction reporting in the Northern Region. Br J Clin Pharmacol. 1992;34:421–426. [PMC free article] [PubMed] [Google Scholar]
- 11.Rawlins MD. Spontaneous reporting of adverse drug reactions II. Uses. Br J Clin Pharmacol. 1988;26:7–11. doi: 10.1111/j.1365-2125.1988.tb03357.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wilson AB. Post-marketing surveillance of adverse reactions to new medicines. Br Med J. 1997;2:1001–1003. doi: 10.1136/bmj.2.6093.1001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Walker SR, Lumley CE. The attitudes of general practitioners to monitoring and reporting adverse drug reactions. Pharmaceut Med. 1986;1:195–203. [Google Scholar]
- 14.Rogers AS, Israel E, Smith CR, et al. Physician knowledge, attitudes and behavior related to reporting adverse drug events. Arch Intern Med. 1988;148:1596–1600. [PubMed] [Google Scholar]
- 15.Fletcher AP. Approaching a century and after: morbidity and potentially drug induced disease. Adverse Drug Reaction Toxicol Reviews. 1995;1:45–64. [PubMed] [Google Scholar]
- 16.Waller PC. Postmarketing surveillance. The viewpoint of a newcomer to pharmacoepidemiology. Drug Information J. 1991;25:181–186. [Google Scholar]
- 17.Inman W, Pearce G. Prescriber profile and post-marketing surveillance. Lancet. 1993;342:658–661. doi: 10.1016/0140-6736(93)91763-c. [DOI] [PubMed] [Google Scholar]
- 18.Waller PC, Coulson RA, Wood SM. Regulatory Pharmacovigilance in the United Kingdom. Current principles and practice. Pharmacoepidemiol Drug Safety. 1996;5:363–375. doi: 10.1002/(SICI)1099-1557(199611)5:6<363::AID-PDS249>3.0.CO;2-7. [DOI] [PubMed] [Google Scholar]
- 19.Anonymous. Pharmacists' adverse drug reaction reporting to start on April 1. Pharm J. 1997;258:330–331. [Google Scholar]
- 20.Davis S, Coulson RA, Wood SM. Adverse drug reaction reporting by hospital pharmacists: the first year. Pharm J. 1999;262:366–367. [Google Scholar]
- 21.Green CF, Mottram DR, Raval D, Proudlove C, Randall C. Community pharmacists' attitudes to adverse drug reaction reporting. Int J Pharm Pract. 1999;7:92–99. [Google Scholar]
- 22.Fisher S, Bryant SG. Postmarketing surveillance of adverse drug reactions: patient self-monitoring. J Am Board Fam Pract. 1992;5:17–25. [PubMed] [Google Scholar]
- 23.Roberts PI, Booth TG, Wolfson DJ. The community pharmacist and adverse drug reaction monitoring: (1) a historical and geographical perspective. Pharm J. 1993;250:875–878. [Google Scholar]
- 24.Campbell JPM, Howie JGR. Involving the patient in reporting adverse drug reactions. J Roy Coll General Pract. 1988;38:370–371. [PMC free article] [PubMed] [Google Scholar]
- 25.Fisher S, Bryant SG, Solovitz BL, et al. Patient-initiated Postmarketing surveillance: a validation study. J Clin Pharmacol. 1987;27:843–854. doi: 10.1002/j.1552-4604.1987.tb05578.x. [DOI] [PubMed] [Google Scholar]
- 26.Fisher S, Bryant SG, Kent TG, et al. Patient drug attributions and postmarketing surveillance. Pharmacother. 1994;14:202–209. [PubMed] [Google Scholar]
- 27.Fisher S, Kent TA, Bryant SG. Postmarketing surveillance by patient self-monitoring: Preliminary data for sertraline versus fluoxetine. J Clin Psychiat. 1995;56:288–296. [PubMed] [Google Scholar]
- 28.Mitchell AS, et al. Adverse drug reactions: Can consumers provide early warning? Pharmacoepidemiol Drug Safety. 1994;3:257–264. [Google Scholar]
- 29.Egberts TC, Smulders M, de Koning FH, et al. Can adverse drug reactions be detected earlier? A comparison of reports by patients and professionals. Br Med J. 1996;313:530–531. doi: 10.1136/bmj.313.7056.530. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.The Royal Pharmaceutical Society of Great Britain. London: the Royal Pharmaceutical Society of Great Britain; 1997. From Compliance to Concordance. Achieving Showed Goals in Medicine Taking. [Google Scholar]
- 31.Corso DM, Pucino F, De Leo JM, et al. Development of a questionnaire for detecting potential adverse drug reactions. Ann Pharmacother. 1992;26:890–896. doi: 10.1177/106002809202600704. [DOI] [PubMed] [Google Scholar]
- 32.Jarernsiripornkul N, Krska J, Capps PAG, Richards RME. Pharmacist-assisted patient-reporting of adverse drug reactions. Pharm J. 1998;261:R33. [Google Scholar]
- 33.Altman DG. 1. London: Chapman & Hall; 1992. Practical Statistics for Medical Research; pp. 403–409. [Google Scholar]
- 34.Mitchell AS, Henry DA, Sanson-Fisher R, O'Connell DL. Patients as a direct source of information on adverse drug reactions. Br Med J. 1988;297:891–893. doi: 10.1136/bmj.297.6653.891. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Willison DJ, Gaebel KA, Borden EK, et al. Experience in the development of a postmarketing surveillance network: the pharmacy medication monitoring program. Ann Pharmacother. 1995;29:1208–1213. doi: 10.1177/106002809502901203. [DOI] [PubMed] [Google Scholar]
- 36.Fisher S, Bryant SG. Postmarketing surveillance. Accuracy of patient drug attribution judgments. Clin Pharmacol Ther. 1990;48:102–107. doi: 10.1038/clpt.1990.123. [DOI] [PubMed] [Google Scholar]
- 37.Rickels K, Downing RW. Side reactions in neurotics. II. Can patients judge which symptoms are caused by their medication? J Clin Pharmacol. 1970;10:298–305. doi: 10.1177/009127007001000502. [DOI] [PubMed] [Google Scholar]
- 38.Sitzia J, North C, Stanley J, et al. Side effects of CHOP in the treatment of non-Hodgkin's lymphoma. Cancer Nursing. 1997;20:430–439. doi: 10.1097/00002820-199712000-00007. [DOI] [PubMed] [Google Scholar]
- 39.Buckingham R, Fitt J, Sitzia J. Patients' experiences of chemotherapy: side-effects of carboplatin in the treatment of carcinoma of the ovary. Eur J Cancer Care. 1997;6:59–71. doi: 10.1111/j.1365-2354.1997.tb00270.x. [DOI] [PubMed] [Google Scholar]
- 40.Brown SW, Tomlinson LL. Anticonvulsant side-effects: a self-report questionnaire for use in community surveys. Br J Clin Pract. 1982;18(Suppl):147–149. [Google Scholar]
- 41.Lee S. Side effects of chronic lithium therapy in Hong Kong Chinese: An ethnopsychiatric perspective. Culture Med Psychiatry. 1993;17:301–320. doi: 10.1007/BF01380006. [DOI] [PubMed] [Google Scholar]
- 42.Wallin J, Sjovall J. Detection of adverse drug reactions in a clinical trial using two types of questioning. Clin Ther. 1981;3:450–452. [PubMed] [Google Scholar]
- 43.Downing RW, Rickels K, Meyers F. Side reactions in neurotics: I. A comparison of two methods of assessment. J Clin Pharmacol. 1970;10:289–297. doi: 10.1177/009127007001000501. [DOI] [PubMed] [Google Scholar]
- 44.Fisher S, Bryant SG, Kluge RM. Detecting adverse drug reactions in postmarketing surveillance: interview validity. Drug Information J. 1987;21:173–183. [Google Scholar]
- 45.Rosenthal J, Bahrmann H, Benkert K, et al. Analysis of adverse effects among patients with essential hypertension receiving an ACE Inhibitor or Beta-Blocker. Cardiology. 1996;87:409–414. doi: 10.1159/000177129. [DOI] [PubMed] [Google Scholar]
- 46.Ciccolunghi SN, Chaudri HA. A methodological study of some factors influencing the reporting of symptoms. J Clin Pharmacol. 1975;15:496–505. doi: 10.1002/j.1552-4604.1975.tb01471.x. [DOI] [PubMed] [Google Scholar]
- 47.Curb JD, Borhani NO, Blaszkowski TP, et al. Patient-perceived side effects to antihypertensive drugs. Am J Prevent Med. 1985;1:36–40. [PubMed] [Google Scholar]
- 48.Fisher S. Patient self-monitoring: challenging approach to pharmacoepidemiology. Pharmacoepidemiol Drug Safety. 1995;4:359–378. [Google Scholar]
- 49.Mannesse CK, Derkx FH, de Ridder MA, et al. Adverse drug reactions in elderly patients as contributing factor for hospital admission: cross sectional study. Br Med J. 1997;315:1057–1058. doi: 10.1136/bmj.315.7115.1057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Scott HD, Rosenbaum SE, Water WJ, et al. Rhode Island physicians' recognition and reporting of adverse drug reactions. Rhode Island Med J. 1987;70:311–316. [PubMed] [Google Scholar]
- 51.Kramer MS, Leventhal JM, Hutchison JA, et al. An algorithm for the operational assessment of adverse drug reactions. JAMA. 1979;242:623–632. [PubMed] [Google Scholar]
- 52.Naranjo CA, Lanctot KL. Microcomputer-assisted Bayesian differential diagnosis of severe adverse reactions to new drugs: 4-year experience. Drug Information J. 1991;25:243–250. [Google Scholar]
- 53.Venulet J. Incomplete information as a limiting factor in causality assessment of adverse drug reactions and its practical consequences. Drug Information J. 1986;20:423–431. [Google Scholar]
- 54.Cipolle RJ, Strand LM, Morley PC. USA: McGraw-Hill; 1998. Pharmaceutical Care Practice. [Google Scholar]
- 55.Krska J, Cromarty JA, Arris F, Jamieson D, Hansford D. Providing pharmaceutical care using a systematic approach. Pharm J. 2000;265:656–660. [Google Scholar]
- 56.Calis KA. Safety reporting in clinical trials. JAMA. 2001;285:16. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
The questionnaire used in this study is available from http://www.blackwell-science.com/products/journals/suppmat/BCP/BCP1547/BCP1547sm.htm