A CONTROLLED EVALUATION OF DIAGNOSTIC CRITERIA IN THE DEVELOPMENT OF A RHEUMATOLOGY EXPERT SYSTEM

As an intermediate stage in the development of an expert systemto support undergraduate teaching in rheumatology, a decisiontree incorporating the diagnostic criteria to be used in theexpert system was produced by a team of rheumatologists. Ina controlled trial, 119 final-year medical students each diagnosed10 rheumatology cases, drawn from a pool of 96 cases, with orwithout the aid of the decision tree. Students who used the decision tree correctly diagnosed thefollowing conditions more frequently than the control group:polymyalgia rheumatica (p<0.05), myopathies (p<0.01),systemic lupus erythematosus (p<0.05), pyrophosphate arthropathy(p<0.05), seronegative spondylarthropathies (p<0.0l),intra articular bleeding (p<0.05) and traumatic synovitis(p<0.05). The overall diagnostic accuracy of the studentswho used the decision tree was 81% compared with 68% for thecontrol group (p<0.001). KEY WORDS: Rheumatology, Decision tree, Diagnosis, Expert system, Medical education     

Factors Affecting the Level of Interest and Activity in Primary Care Research Among General Practitioners

This paper presents the results of a state-wide survey of generalpractitioners in Victoria, Australia. The survey assessed theirlevel of interest in general practice research as well as factorsinfluencing the decision to participate in research. The studywas conducted during the formation of a general practice researchnetwork which was being linked to a new primary care researchunit in the Department of Community Medicine at Monash University.Of the 3350 general practitioners surveyed 1116 responded. Onethird expressed a high degree of interest in general practiceresearch. Previous research experience and barriers to generalpractice research were also studied. The areas in which thedoctors felt most research was required were identified. Theresults are compared with another study of Canadian family physicians,which is the only study of a similar nature previously published.The implications of the results are discussed with referenceto the future direction of research by general practitioners.     

Associations of smoking with prescribed medications, arthritis/rheumatism and back problems in the British Health and Lifestyle Survey follow-up respondents

The relationship of smoking behaviour to the prevalence of prescribedanalgesics and/or anti-inflammatory agents, reported treatedarthritis/rheumatism or back problems has been examined. The5, 352 survivors, now aged 25–99 years, of the 1984–1985nationwide Health and Lifestyle Survey of randomly selectedBritish adults were re-interviewed in 1991–1992. Therewere 2, 300 men and 3, 052 women. The response rate was 80.8%of those traced and alive. Of those interviewed 4, 483 subjects(83.8%) were also visited by a nurse. Details of prescribedmedications and the reasons for taking them were collected bythe nurses and information on treated medical conditions wasbased on the responses during the interview. Age and socioeconomicgroup were controlled for in the analyses. There were positiveassociations, for the sexes combined, of a smoking history withcurrent prescribed analgesics (p<0.02), non-steroidal anti-inflammatoryagents (NSAIDS) (p<0.05) and analgesics and NSAIDS combined(p<0.005). Arthritis/rheumatism was more often reported bycurrently smoking men under 65 years (p<0.05) and more noticeablywomen under 55 years with a smoking history (p<0.001). Thepreviously recognized association between back problems andsmoking has been confirmed with current and past smokers, inparticular women, being more likely to be sufferers (p<0.001).In subjects who reported arthritis/rheumatism, those with asmoking history were also more likely to be currently takingthe medications than non-smokers (ORs of 1.90 and 1.67 for menand women respectively).     

The Mammography Controversy A Case for Breast Self-Examination

A plea for renewed emphasis on the breast self-examination is included with a review of the recent controversy over risks vs. benefits in mammographic screening for breast cancer. A discussion of the implication for nursing practice is presented, and facts patients should know about mammography are listed.     

Juvenile Diabetes Mellitus—a Risk Factor for Postoperative Venous Thromboembolism?

ABSTRACT In a retrospective analysis of 125 patients and a prospective evaluation of 83 patients with terminal uremia undergoing kidney transplantation, juvenile diabetes mellitus was found to be a significant risk factor for the development of postoperative thromboembolism. We found a high frequency of objectively verified thromboembolism despite the relatively young age of the patients. Besides diabetes, no other clinical risk factor differed between patients with and without thrombosis.     

Exercise induced asthma: response to disodium cromoglycate in skin-test positive and skin-test negative subjects

Thirty-nine asthmatic subjects, aged 5-50 and each with a history of exercise-induced asthma, were classified according to their skin response to prick tests using nineteen common antigens. Ten had negative skin tests, four responded only to D. farinae and twenty-five had multiple positive responses. Each patient then carried out three exercise tests on a treadmill, each test on a separate day. A control test was followed, in random order, by an exercise test after administration of disodium cromoglycate or of a placebo. In all groups, the mean fall in peak expiratory flow rate was less after disodium cromoglycate than after placebo, but the difference was significant only for the skin-test positive groups. Similarly, positive skin-test groups had a higher incidence of drug responders than did the negative skin-test group. These observations are discussed.     

A double-blind trial of clomocycline in the treatment of persistent palmoplantar pustulosis

Sixty patients with persistent palmoplantar pustulosis (PPP) were treated with clomocycline (Megaclor) in a double-blind, cross-over trial in an attempt to establish whether the condition responds to tetracycline. Each patient received 3 months each of clomocycline and placebo in random order. Forty patients completed the trial. Twenty-two failed to respond to either treatment, fifteen improved on clomocycline, two improved on placebo and one improved on both treatments. These results are highly significant (P= 0-003) d suggest that clomocycline may suppress pustulation in some patients. The twenty-two‘non-responders' were compared with the eighteen‘responders' for sex, age, length of history and associated psoriasis but no significant differences were found. Difficulties in assessment and the possibility of improvement being due to spontaneous remission are discussed. Further follow-up of both groups suggested that clomocycline used over long periods favourably influenced the course of the disease in the‘responder' group.     

Immunoreactive' TSH' in urinary concentrates of Graves' disease patients: a radioimrnunoassay artefact

Abstract. A double antibody radioimmunoassay was used to analyse immunoreactive thyrotrophin in urinary concentrates from fourteen patients with hyperthyroid-ism due to Graves' disease, in three subjects with primary hypothyroidism, and in six normal subjects. Immunoreactive thyrotrophin was detectable in eleven subjects with Graves' disease, in one subject with primary hypothyroidism, and in four normal subjects. The mean urinary thyrotrophin concentration was significantly higher in Graves' disease (492 ± 99.9 μU/24h (SEM) ( n = 11)) than in normal subjects ((177 ± 26.4 μU/ 24 h (SEM) ( n = 4) (P<0.01)). The fact that bioassay-able thyrotrophin was undetectable in urinary concentrates in Graves' disease and that radioimmunoassay of other polypeptide hormones in urine can be influenced nonspecifically, led to the suspicion that the immunoreactive thyrotrophin measured in urinary concentrates was an assay artefact. This was strongly supported by the absence of a radioimmunoassayable thyrotrophin in urinary concentrates after extensive dialysis (96 h) without loss of added ¹³¹I-labelled thyrotrophin, differences between the elution pattern on Sephadex G-100 columns of ¹³¹I-labelled thyrotrophin and the immunoreactive thyrotrophin detected after chromato-graphy of the urinary concentrates on the same column, and the ability of different salts in various concentration to mimic dose-response curves in the radioimmunoassay for thyrotrophin as well as for thyroglobulin. In addition, since urinary proteins and IgG levels were elevated in concentrates of Graves' disease when compared with normal subjects, it has been suggested that the filtering mechanism of kidney is affected in Graves' disease, a disorder which has recently been associated with circulating antigen—antibody complexes.