Prescription-event monitoring: methodology and recent progress.

Event monitoring was first suggested 25 years ago as a way of detecting adverse reactions to drugs. Prescription-event monitoring (PEM), which has been developed by the Drug Safety Research Unit, is the first large-scale systematic post-marketing surveillance method to use event monitoring in the U.K. PEM identifies patients, who have been prescribed a particular drug, and their doctors from photocopies of National Health Service prescriptions which are processed centrally in England. A personalized follow-up questionnaire ("green form") is mailed to each patient's general practitioner, usually on the first anniversary of the initial prescription, asking for information about the patient, especially any "events" that he or she may have experienced since beginning treatment with the drug. The methodology of PEM is presented, together with examples of analyses that can be performed using results from recent studies. The problems and benefits of PEM are discussed.

Postmarketing surveillance of enalapril. I: Results of prescription-event monitoring.

To identify and measure the incidence of adverse effects of the angiotensin converting enzyme inhibitor enalapril 13,713 patients were studied for one year by prescription-event monitoring. Precise information about the duration of treatment was available for 12,543 patients. The frequency of many events was calculated, including dizziness (483 patients; 3.9%), persistent dry cough (360; 2.9%), headache (310; 2.5%) hypotension (218; 1.7%), and syncope (155; 1.2%). Less common reactions included angioedema, urticaria, and muscle cramps. Altogether 1098 (8%) patients died and the notes of 913 of them (83%) were obtained for detailed scrutiny. With the exception of a few patients with renal failure who deteriorated during treatment (reported on separately), no death was attributed to enalapril. Enalapril was considered to be effective, even in patients with advanced cardiac failure. These results for enalapril are reassuring and provide further evidence of the value of prescription-event monitoring.

Postmarketing surveillance of enalapril. II: Investigation of the potential role of enalapril in deaths with renal failure.

The possibility that enalapril might damage renal function was investigated in 1098 deaths recorded in a prescription-event monitoring study. Case notes for 913 patients were examined. In seventy five there was a rise in the urea or creatinine concentration of 50% or more above pretreatment values. Enalapril appeared to have contributed to a decline in renal function and subsequent death in 10 of these patients. Several characteristics were identified among these patients, including old age, the use of high dose or potassium sparing diuretics, and pre-existing renal disease. Adding a non-steroidal anti-inflammatory drug was also associated with a deterioration in patients with previously stable renal function. No death was encountered of a patient with uncomplicated hypertension. Enalapril infrequently contributed to a substantial decline in renal function in certain vulnerable patients, especially those receiving other drugs known to be capable of adversely affecting renal function. Awareness of the characteristics of these patients and of their concomitant treatment may serve to reduce the risk.

Prescription-event monitoring. A preliminary study of benoxaprofen and fenbufen.

Prescription-Event Monitoring (PEM) has been established at the Drug Surveillance Research Unit of the University of Southampton as a low-cost technique for ascertaining the pattern of events, whether drug-related or not, in large general practice cohorts. The reporting of "events" without the need for an opinion about the probability that they may be adverse drug reactions (ADRs) removes much of the uncertainty inherent in voluntary ADR reporting systems. Numerators (adverse events) and denominators (the number of prescriptions), enable estimates of incidence to be derived from the data. Where related drugs are studied concurrently, differences in the pattern of events may signal important differences in their safety or efficacy . A successful large-scale preliminary exercise involving nearly 9 000 doctors and 16 000 patients is described.

Blood disorders and suicide in patients taking mianserin or amitriptyline.

26,781 patients who had been prescribed mianserin and 42,082 prescribed amitriptyline were followed up by means of a questionnaire addressed to their general practitioners. No patient had aplastic anaemia, agranulocytosis, or leucopenia severe enough to endanger life, and in only 2 patients in each group was a causal association likely. If either drug dose cause leucopenia, the incidence is likely to be in the range of 1 in 10,000 to 1 in 100,000 patients. Among patients who were reported to have attempted suicide, 56 of 246 survivors of amitriptyline overdoses required intensive care, compared with none of 92 patients who overdosed with mianserin. 4 patients who overdosed with amitriptyline alone died, compared with none of the patients who overdosed with mianserin alone. Both drugs are associated with a low risk of blood disorders, but mianserin is appreciably safer than amitriptyline because of its low toxicity in overdosage.

Study of fatal bone marrow depression with special reference to phenylbutazone and oxyphenbutazone.

The histories of 269 patients whose death certificates did not mention a drug as the cause of aplastic anaemia or agranulocytosis were investigated. Eighty-three deaths were probably caused by drugs, the most common cause of aplastic anaemia being treatment with phenylbutazone (28 deaths) and oxyphenbutazone (11 deaths). Thirteen out of 17 deaths from agranulocytosis were attributed to co-trimoxazole treatment. A separate survey of general practitioners' prescriptions enabled the mortality to be estimated. With the addition of one death due to oxyphenbutazone and four deaths due to phenylbutazone that were reported independently to the committee, the mortality from oxyphenbutazone was 3-8 per 100 000 and from phenylbutazone 2-2 per 100 000. With phenylbutazone the rates varied from under 1 death per 100 000 for men aged under 65 years to 6 per 100 000 for women aged 65 and over. Small numbers precluded estimates for oxyphenbutazone in these subgroups, although a similar trend was suggested. No particular indication for treatment seems to carry a higher risk, the main concern being the use of these two drugs in elderly patients.

Postmarketing surveillance of adverse drug reactions in general practice. II: Prescription-event monitoring at the University of Southampton.

An independent, non-regulatory drug surveillance research unit has been established at the University of Southampton. Its first task will be to set up a prescription-event monitoring scheme in general practice to enable the pattern of adverse events, as distinct from suspected adverse reactions associated with new drugs to be compared with that of older medicines. Prescriptions for selected drugs will identify patients and a simple questionnaire, designed to be completed in under five minutes, will be used to obtain the required information. Medical opinions about causation need not be given, and the scheme will not interfere with normal prescribing practice.

Oral contraceptives and fatal subarachnoid haemorrhage.

A case-control study was conducted of the deaths from subarachnoid haemorrhage (SAH) in women aged 15-44 in England and Wales in 1976. There was a small excess of oral contraceptive use by the women who died from SAH compared with their generally healthy practice-matched controls; this was not, however, statistically significant. Out of 134 women who died from SAH, 34 had a history of hypertension compared with only six of their controls. Renal disease and pre-eclamptic toxaemia were more commonly associated with hypertension in the dead women than in controls. No change in the annual mortality from SAH has been observed in the past 20 years such as might have been expected if the risks were high. Although current or past use of oral contraceptives may have increased the blood pressure and risk of SAH in a few women, the most important factor in determining this risk was hypertension. SAH should thus probably not be regarded as serious cause for concern in healthy non-hypertensive women using oral contraceptives.

Postmarketing surveillance of adverse drug reactions in general practice. I: search for new methods.

Spontaneous reporting of suspected adverse drug reactions through the yellow card system provides an effective means of early warning of potentially serious drug effects but occasionally fails because of under-reporting. Since more than 80% of medicines are prescribed by general practitioners we look to them to support new methods of postmarketing surveillance. No single method will be appropriate for all drugs, and yellow cards will always be required to detect rare events. It is proposed that adverse events experienced by the first 10 000 patients who receive certain new medicines should be recorded and transmitted to a monitoring centre and that a realistic target, which should be reached early in the market life of a new medicine, is the detection of any adverse event that occurs in more than in 1000 patients.

Non-steroidal anti-inflammatory drugs: assessment of risks.

The prescription-event monitoring procedure developed at the University of Southampton was used to evaluate five different non-steroidal anti-inflammatory drugs (NSAIDs), including piroxicam, in approximately 55,000 patients. The overall incidence of side effects was what would be expected. The risks of gastrointestinal haemorrhage and peptic ulceration were spread uniformly across the five drugs under study. No real difference was seen in the incidence of these conditions when patients discontinued the medication or switched to another NSAID. Overall, serious side effects were extremely rare. One provocative finding was the possibility that drugs of this class may exert a cardioprotective effect. There appeared to be a deficit of cases of myocardial infarction while the patients were taking an NSAID. Responsibility for the efficacy and safety of all drugs resides with many people and organizations: the manufacturers, national health departments, licensing departments, and physicians. The press, which has contributed to many unwarranted panics concerning drug safety, also must refrain from using power without responsibility.

Stimulation of glucose uptake by transforming growth factor beta: evidence for the requirement of epidermal growth factor-receptor activation.

Transforming growth factor beta (TGF-beta), derived from human platelets, stimulates the uptake of 2-deoxy-glucose by cultured cell monolayers 2- to 4-fold. Stimulation can be detected as early as 30 min with as little as 0.1 ng of TGF-beta per ml and maximal effects can be obtained at 2 hr with 1 ng of the growth factor per ml. TGF-beta-induced stimulation of sugar uptake is enhanced by the co-addition of platelet-derived growth factor (10 ng/ml) or epidermal growth factor (EGF, 1 ng/ml). The NR-6 variant of mouse 3T3 cells, which lack EGF receptors, is not stimulated by TGF-beta. Antisera to EGF receptors that block 125I-labeled EGF binding also inhibit TGF-beta stimulation of 2-deoxyglucose uptake, although 125I-labeled TGF-beta binding remains unimpaired. In contrast, anti-sera to the EGF receptor, which do not block EGF binding, have no measurable effect on the TGF-beta-stimulated uptake of 2-deoxyglucose. We confirm that the receptor for TGF-beta is distinct from the receptor for EGF and we conclude that TGF-beta stimulation of 2-deoxyglucose uptake requires the co-activation of the EGF receptor kinase system.

Oral contraceptives and death from myocardial infarction.

We investigated 219 deaths from myocardial infarction in women under the age of 50. Their histories were compared with those of living age-matched controls selected from the same general practices. The frequency of use of oral contraceptives during the month before death was significantly greater in the group with infarction than during the corresponding month in the control group and the average duration of use was longer. No information of cigarette smoking was available but the proportion of women being treated for hypertension or diabetes was greater among those who died than among the controls. This did not alter the overall conclusion that the risk of fatal myocardial infarction was greater in the women using oral contraceptives, particularly in the older age groups.