A prospective multipractice study based on antecedent data. Reduced selection and behaviour bias

A prospective multipractice study based on well-defined data collected by a number of general practitioners (GPs) over a period of time is often used in primary health care research. However, problems of research methodology such as selection bias are common in such studies when GPs either refuse to participate or interrupt their participation. A method is described whereby selection bias is avoided. In the present study, a medical laboratory made a daily record of all abnormal liver function tests requested by GPs over a period of six months. When mailing these results to the GPs, a questionnaire was enclosed in order to obtain a number of clinical data from them. Thirty of 31 possible GPs participated. During the study period, questionnaires concerning 134 patients were mailed, and 118 of these were returned without reminders. Following one reminder the response rate was 100%. Comparing the number of requested tests during the study period with the number requested during a control period before and after the study period (same six month periods of the year), no difference was seen in the pattern of requests. Thus, this method, unbiased with respect to both selection and request-behaviour, is recommended for similar studies.

Epidemiology of abnormal liver function tests in general practice in a defined population in Denmark

It is common for general practitioners (GPs) to refer patients suspected of impaired liver function for laboratory tests (alkaline phosphatase, lactate dehydrogenase, bilirubin, prothrombin, aspartate aminotransferase). In a prospective multipractice study over a six-month period, including 30 GPs, 55 patients were recorded as having, for the first time, a high level of alkaline phosphatase (AP) as an isolated finding, 14 with an increase of aspartate aminotransferase (ASAT), eight with an increase of both AP and ASAT, three with an increase of ASAT, AP, and bilirubin, two with an isolated increase of lactate dehydrogenase (LDH), one with an increase of ASAT, AP, and bilirubin, combined with a low prothrombin (PP), and, finally, one patient with a low prothrombin in isolation. In most cases the tests were requested because of unspecific symptoms. The most common causes of abnormal test results were neoplasms, alcoholic liver disease, and heart failure. Thirty patients were referred to hospital for further investigations. During the same study period, 50 patients with known abnormal liver function tests were recorded, and the most common causes of these abnormalities were neoplasms, rheumatoid arthritis, and alcoholic liver disease.

Reduced occurrence of atrial fibrillation in acute myocardial infarction treated with streptokinase

In a historical follow-up study of 152 hospital patients with acute myocardial infarction, the frequency of life-threatening arrhythmias (ventricular fibrillation, sustained ventricular tachycardia, 3rd degree AV-block, 2nd degree AV-block (Mobitz type II), and asystole) and atrial fibrillation in 76 patients treated with streptokinase was compared with their frequency in 76 patients who did not receive a thrombolytic therapy. Among those treated with streptokinase two patients (3%) developed atrial fibrillation, compared with 12 (16%) in the control group (P = 0.009). Life-threatening arrhythmias occurred with equal frequency in the two groups. Further studies should confirm and clarify the mechanism of the reduced frequency of atrial fibrillation in the streptokinase-treated patients.

[Epicrises from a department of medical gastroenterology]

The content of discharge letters is important to general practice in order to give an optimal physical, psychiatric and social treatment in view of ideal utilization of the resources in the primary health service. The content of 91 discharge letters from a medical department of gastroenterology was judged by the general practitioner. Most often medication on discharge (25%) and reason for referral (15%) were missing. Dissatisfaction was expressed about late receipt of discharge letters (median: 13 days), the lack of information given to the patients, medicine on discharge, as well as reasons for possible changes in medication. The quality and content of the discharge letter were evaluated as good on the whole (80%). It is concluded that the discharge letter should have a higher priority and, for example, training is necessary. Further investigations are needed to solve these problems.

[Diagnostic problems of purulent meningitis in children in general practice]

Previous investigations have demonstrated that considerable problems may exist in the diagnosis of purulent meningitis (PM) in general practice. Referrals from general practitioners/duty roster doctors concerning 97 children discharged the diagnosis of PM were reviewed retrospectively. The patients were subdivided into two groups according to whether the diagnosis was established by the referring doctor or not. Only 35% of the children under one year were admitted for suspected PM, whereas 65% of the children between one year and 15 year were hospitalized with the correct diagnosis. The commonest positive findings in both age groups were alterations in consciousness which were found in more than 80% of the children. Children in whom the diagnosis was not established by the referring doctor had fewer classical signs of meningitis (neck-stiffness, Kernig's sign, bulging fontanelle and petecchia) than children in whom the diagnosis was established. Children with negative cultures from the cerebro-spinal fluid were significantly more frequently treated with antibiotics prior to hospitalization. Approximately half of the children admitted with suspected meningitis were not treated according the guidelines issued by the Danish Board of Health, without this having any effect on the survival rate.

Lauge-Hansen classification of malleolar fractures. An assessment of the reproducibility in 118 cases

The radiographs of malleolar fractures in 118 consecutive patients were assessed by 4 observers using the Lauge-Hansen classification. The interobserver variation was large; only 51 of the 118 radiographs were classified identically by all 4 observers. With the purpose of assessing intraobserver variation, the same observers classified the same radiographs 6 weeks later. The first and the second classification were identical in 75-97 of 118 radiographs (64-82 percent). These results show that the Lauge-Hansen system is difficult to apply in a reproducible way.

[Referral to a department of medical gastroenterology. A study of the content of the referral letter and the reasons for referral]

Knowledge concerning factors that affect referrals of patients by general practitioners to hospitals is important in view of optimal utilization of the resources. Out of 108 referrals from general practitioners to a medical department of gastroenterology, information about the medical history was insufficient in about half the cases, and in 3/4 of the cases information about medicine was absent. Information about allergy and social conditions was found in only a few cases. The reason for referral was in most cases to establish a diagnosis and treatment. It is concluded, that if functions in a medical department of gastroenterology are to be moved to the general practitioners, more education in diagnostic procedures and treatment is needed. Further investigations are needed to solve these problems.

[Reproducibility of urine strip reactions in medical patients]

Before a method of investigation is introduced in the daily clinical routine, its reproducibility and correctness must be proved. In the present investigation, the reproducibility of urine stix reactions for glucose, haemoglobin and protein were assessed. The urine stix tests were carried out on freshly voided urine from 30 patients selected at random by three junior nurses with seniorities of one, five and nine years. Complete inter- and intraobserver agreement was found for glukose and haemoglobin whereas the reproducibility for protein was unsatisfactory.

[Diagnosis of penicillin allergy. Results of a clinical investigation program]

Investigations concerning penicillin allergy (PA) from abroad suggest that considerable overdiagnosing occurs. The absence of standardized commercially available preparations for skin testing and the time necessary for thorough testing are also contributory causes. Fifty patients with suspected PA were investigated with a programme consisting, in serial order, of RAST for IgE to penicillin V and G, a cutaneous test with a commercially available test preparation (Penkit), first with a prick followed by intracutaneous testing and finally with oral provocation. If a positive reaction occurred, the test was discontinued. Six patients (12%) had positive RAST or cutaneous reactions, (1 positive to RAST, 2 to prick and three not until the intracutaneous test), while no patients with negative RAST and cutaneous tests reacted to oral provocation. No generalized allergic reactions occurred during testing. It did not prove possible to predict which patients would react positively on the basis of the case histories. This investigation thus demonstrates that PA is also considerably overdiagnosed in Denmark and that a programme of investigation such as this is suitable for testing for PA and, finally, that the programme of investigation can scarcely be simplified.

[Attitudes to the use of vaginal cytologic examinations in general practice]

In June 1988, a questionnaire was sent to 221 Danish general practitioners chosen at random and to 195 registrars who had applied for postgraduate courses in general medicine. The percentage replies were 83 and 73, respectively. Employing the questionnaire, information was obtained about the doctor's employment of CS on the basis of definite case reports and the doctor's assessment of the certainty of the examination. The investigation revealed that there was a general tendency to freer indications than those officially recommended, that younger women were examined more frequently than recommended, that elderly patients were examined less frequently than recommended and that considerable excessive employment of CS took place in women with condylomata and herpes genitalis. As a rule, the doctors overestimated the reduction in the risk of disease obtained by CS. It is concluded that information from the official authorities concerning employment of CS sent to general practitioners is not satisfactory and that this should be considered in extension of future guidelines about the employment of methods of examination. Precise and effectively distributed information about rationel employment of an examination must be considered to be an important way in which to obtain optimal utilization of resources.

[A study of cardiac vagal function in chronic alcoholics]

Reduced cardiac vagal tonus is correlated to sudden cardiac death and occurrence of arrhythmia, particularly in patients with ischemic heart disease. In patients with diabetic neuropathy increased frequencies of silent myocardial ischaemia and silent myocardial infarctions have been found, in addition. The object of the present investigation was to assess whether chronic alcoholics have lowered cardiac vagal tonus. The cardiac vagal tonus was assessed by means of a sensitive non-invasive method in which the variations in the pulse rate during six deep respirations were registered on the basis of the R to R interval in the ECG. Fourteen alcoholics and fourteen healthy controlled individuals participated in the investigation. The medians of the variations in the two groups were 15 and 25 beats per minute, respectively. The difference was statistically significant. No significant differences were observed in alcoholics with or without clinical evidence of polyneuropathy. This investigation thus suggests that ethanol has a toxic effect on the cardiac vagal function.

[Pain in acute myocardial infarction]

Pain is the most constant symptom in acute myocardial infarction (AMI) but there are considerable variations. It is possible that pain may indirectly cause extension of the infarct. The authors have, therefore, undertaken a retrospective investigation of 87 unselected patients admitted consecutively with the first episode of AMI. In these patients, the presence of any clinical and paraclinical parameters which could predict the employment of analgesics and the duration of pain while hospitalized were investigated. A significant connection was found between the presence of pulmonary stasis on administration to hospital and the total employment of analgesics during hospitalization. Patients with pulmonary stasis have frequently extensive infarcts and a significant connection was found between the enzyme-estimated extent of the infarct and the employment of analgesics and duration of pain. In contrast to previous investigations which were carried out on selected patient materials with exclusion of large infarcts, no significant connection was found between heart rate, systolic blood-pressure, electrocardiographic signs of AMI on admission and total amount of analgesics employed and the duration of pain.

The relationship between serum enzyme activity, infarct site, and cardiac complications after a first myocardial infarction. A follow-up study in general practice

The aim of the present study was to relate the clinical course in patients after a first acute myocardial infarction with the site and size of infarct, estimated from standard enzyme measurements. One hundred and eight consecutive patients who suffered an acute myocardial infarction for the first time were followed-up after 30 months in general practice. Twenty-six patients had died and 8 had had another infarction. Sixty-two of the surviving patients had received treatment for ischaemic heart disease, usually for angina pectoris and less often for heart failure and arrhythmias. No correlation was found between ischaemic heart disease requiring treatment and the enzyme-estimated size or the site of the infarct. With anterior infarcts there was, however, an overweight of arrhythmias requiring treatment. Of the patients at work, 31% had changed job or job status because of ischaemic heart disease. At the end of the 30 month period, 50 patients were in functional class 1 and 2, and 32 in functional class 3 and 4 (New York Heart Association's classification).

Anaphylactic shock occurring outside hospitals

In a retrospective study, all cases of anaphylactic shock (AS) occurring in a hospital catchment area during a 13-year period were analysed. Twenty cases were found, giving an incidence of 3.2 cases per 100,000 inhabitants per year, and one patient died. The precipitating agents were penicillin (7 cases), aspirin (3), food (2), and bee or wasp sting (8). The drugs had all been given orally. Penicillin-related AS was much commoner than its incidence reported by the Danish national health authorities. In three cases, previous anaphylactic reactions had occurred to the agent in question. We concluded that AS rarely occurs outside hospital, but that oral penicillin is a more important cause than was previously thought. Many recurring episodes of AS should be preventable, since hypersensitivity tests for both penicillin and hymenoptera venoms are available, and hyposensitization with respect to the latter is feasible.

Differences in referral rates from general practice

There are many unexplained differences in the rates at which general practitioners make referrals to other medical specialists. This study investigated 17,586 referrals from 141 general practitioners to specialists in seven specialties in Ringkjøbing county in Denmark. As an expression of the referral rate, a referral index was estimated for every general practitioner. The referral index was the number of referrals to the specialist per 1000 patients per year, including children, standardized for age and sex to the average population in Ringkjøbing county. The following six variables were evaluated in relation to the referral index: specialists in the local area, doctors per practice, consultations per general practitioner per year, patients registered, consultations per 1000 patients per year standardized for age and sex, and supplementary procedures per consultation. Stepwise multiple regression analysis was used. The study showed that the referral index rose both with a better access to specialist and with an increasing number of consultations per practitioner per year. The referral index fell with increased numbers of patients registered. No correlation was found between the referral index and number of supplementary procedures per consultation, number of doctors per practice and number of consultations per 1000 patients per year.