Value of screening for secondary causes of hyperlipidaemia in general practice
BMJ 1994; 309 doi: https://doi.org/10.1136/bmj.309.6953.509 (Published 20 August 1994) Cite this as: BMJ 1994;309:509
- P Evans,
- D P Gray
- Correspondence to: Dr Evans.
- Accepted 14 April 1994
The number of estimations of serum cholesterol concentrations being performed has risen dramatically in the past few years. The reason for the rise is partly that general practitioners are increasingly realising the importance of cholesterol and partly that authoritative reports - notably that of the Royal College of General Practitioners1 and the European Atherosclerosis Society2 - have encouraged general practitioners to screen for raised cholesterol concentrations.
Leading lipidologists and all the major textbooks emphasise the association between a raised serum cholesterol concentration and diabetes, hypothyroidism, excessive alcohol consumption, and renal failure,3, 4 and virtually all authorities advise general practitioners to make sure that when hyperlipidaemia is present no treatable cause exists. All patients with hyperlipidaemia should therefore have at least blood tests for glucose, thyroid function, renal function, and liver function. These recommendations, however, have been based on theoretical considerations, and the yield of new diagnoses when these tests are done together in primary care has not been reported.
Patients, methods, and results
We identified 330 out of 2630 patients aged >=15 on our personal lists in general practice5 as having a raised serum cholesterol concentration. A raised concentration was defined as a random or fasting cholesterol concentration of >=6.5 mmol/l once during a period of 10 years ending in June 1991. These patients were detected opportunistically during normal surgery consultations according to practice protocols.
The age-sex distribution of the patients on both lists was similar to the national average, and the practice protocol advocated screening patients with established cardiovascular disease, family history, or clinical signs of hyperlipidaemia, obesity, diabetes, or hypothyroidism. The clinical records were then audited to determine whether each patient had been investigated for a secondary cause of their hyperlipidaemia with tests to determine concentrations of random blood glucose; serum thyroxine and thyroid stimulating hormone; and serum creatinine and (gamma)- glutamyltransferase. Patients whose results were abnormal were either managed accordingly or tested again if the diagnosis was uncertain.
Of the 330 patients with a raised serum cholesterol concentration (range 6.5 to 10.6 mmol/l; mean 7.4) 156 were male. The patients were mainly in the older age groups with 127 aged >65 and 50 aged >75 (range 23 to 87). Strenuous efforts were made to achieve complete coverage with each of the screening tests but by July 1993 we had achieved rates of only 86.6-95.5%. The full set of tests, however, was performed in 79% (262) of the patients.
The table shows the results of the tests. The 1542 screening tests yielded six new definitive diagnoses based solely on the screening tests: subclinical hypothyroidism (1); newly diagnosed diabetes mellitus (2); and raised (gamma)-glutamyltransferase (3, who were also dependent on alcohol). Two further patients who had persistently raised concentrations of (gamma)-glutamyltransferase needed further investigation. Eight patients had slightly raised serum creatinine concentrations, four of whom may need further investigation.
Results of 1542 screening tests for secondary hyperlipidaemia in 330 patients with raised serum cholesterol.* Values are numbers of patients
Comment
The total yield of new diagnoses reached on the basis of this series of blood tests was only 6 (1.8% of the patients). At 1993-4 prices this exercise would have cost about pounds sterling 5000 for the tests alone and probably at least as much again for staff time in our practice. We therefore question the place of such biochemical screening for causes of secondary hyperlipidaemia in primary care.
We thank the biochemistry department, Royal Devon and Exeter Hospital (Wonford), for performing the estimations. We thank Drs Russell Steele and Kieran Sweeney for their encouragement and support, Mrs Vi Chadwick and Mrs Christine Pike, and Mrs Sue Woodgate for her clerical and database work. PE received a Royal College of General Practitioners' research training fellowship throughout the study.