Transferring the costs of expensive treatments from secondary to primary care.

General practitioners, especially fundholders, are becoming increasingly concerned about being asked to prescribe treatments for their patients that are outside their therapeutic experience. They are concerned about the clinical responsibility for such prescribing and the effects on their budgets. In some specialties transferring the costs of expensive treatments from secondary to primary care (cost shifting) has become partly institutionalised because of the separate sources of funding for drugs prescribed in the two sectors. With increased efforts to control the rising costs of the drugs budget and the emergence of new expensive treatments, cost shifting will be a challenge to clinicians and purchasers as they strive for rational, cost effective prescribing. A review of the funding mechanisms for drugs prescribing and of the relation between the licensing process and the decision to support the use of a treatment in primary or secondary care is needed.

Population screening for abdominal aortic aneurysm: do the benefits outweigh the costs?

To prevent the high mortality rate associated with ruptured abdominal aortic aneurysm (AAA), population screening at or soon after retirement age has been advocated, with elective operations being performed on patients with the appropriate indications. There is considerable pressure on some health authorities to fund such programmes even though there is substantial uncertainty about the consequent benefits. The ultrasound screening test is acceptable and accurate. Also, other health problems may be detected in the same screen. However, screening would lead to a questionable increase in surgery, as most patients with AAA die from other causes and not from a ruptured aneurysm. In addition, there is an elective operative mortality of around 5 per cent. Furthermore, as many of those who have a positive result on screening would never have known that they harbour an aneurysm, there is the possibility of unnecessary anxiety arising from the test. An economic analysis has been conducted for two identical, hypothetical cohorts of men using the best available data. One cohort was assumed to undergo screening and a number of men were indicated for immediate elective aneurysm repair or for follow-up, and surgery if their aneurysms become large. The other cohort was assumed not to be screened and would thereby face the possibility of rupture with its adverse outcome. The survival prospects of the two cohorts were calculated as life expectancies and in terms of life-years: the incremental life-years gained were compared with the incremental costs of the programme. Although there are considerable uncertainties in the analysis parameters, the base-line result and sensitivity analysis indicate that, on the basis of current knowledge, population screening should not be introduced.

The influence of needle size on post-myelography headache: a controlled trial.

This study aimed to test the hypothesis that the use of fine rather than larger calibre needles for myelography reduces the incidence of post-procedural headache. 160 patients were randomized in a prospective controlled trial of 22 gauge versus 26 gauge needles. The incidence and overall severity of headache as measured at 1, 4 and 21 days after myelography was not significantly different in the two groups.

Further evidence of lipid peroxidation in post-enteropathic haemolytic-uraemic syndrome.

Lipid peroxidation may play a role in the pathogenesis of the haemolytic-uraemic syndrome (HUS). Thirteen children with the post-enteropathic form of HUS were studied using conjugated diene lipids as markers of in vivo lipid peroxidation. Levels of total conjugated diene lipids and 9,11-linoleic acid, the principal conjugated diene in human plasma, were greater in the acute phase of this disorder than in controls. The ratio of plasma vitamin E to lipid was lower than that in children with other renal diseases, and the expected positive correlation between vitamin E and lipids did not hold for HUS patients. These data provide further evidence of lipid peroxidation in HUS and a disturbance in the metabolism of the principal lipid-bound anti-oxidant vitamin E.

Fundholding in general practice and financial risk.

OBJECTIVE: To estimate the financial effect of random yearly variations in need for services on fundholding practices with various list sizes. DESIGN: A simulation model was derived using historical data on general practitioner referrals for the 113 surgical procedures covered by the general practitioner fund, combined with data on the hospital prices for those procedures. PATIENTS: Resident population of Central Birmingham Health Authority. MAIN OUTCOME MEASURES: Expected expenditure on the relevant surgical procedures for the whole district and for practices with list sizes of 9000, 12,000, 15,000, 18,000, 21,000, or 24,000 for each of 100 simulated years. RESULTS: By using average hospital prices for the West Midlands region the mean (SD) annual expenditure for the 179,400 residents was 4,832,471 pounds (87,149 pounds); the random variation between the 5th and 95th most expensive years was 5.7% of the mean cost. For a practice with a list size of 9000 the values were 244,891 pounds (18,349 pounds), with a variation of 27.5%. With a list size of 24,000 the values were 652,762 pounds (32,512 pounds), with a variation of 15.3%. CONCLUSIONS: Random variations in need for inpatient services will have a significant financial impact on the practice fund. The problem will be particularly great for smaller practices. Additional measures are required to ensure that the scheme is not undermined and that the potential benefits are secured.

How long should patients with Salmonella typhi or Salmonella paratyphi be followed-up? A comparison of published guidelines.

Records were examined for 242 individuals infected with Salmonella typhi or S. paratyphi identified in Birmingham between 1981 and 1988, with a total of 335 person years of follow-up. Of these cases 77 and 78 per cent respectively were followed beyond the point at which surveillance would have ceased under guidelines published by the American Public Health Association and by the Public Health Laboratory Service for England and Wales. Under these two sets of guidelines only seven (3.8 per cent) and eight (4.3 per cent) cases respectively had subsequent positive faecal or urine cultures over a median of 335 and 295 days of additional follow-up. After 0, 1, 2, 3, 4 and 5 prior consecutive negative sets of cultures obtained at weekly intervals the likelihood of the next set of cultures being positive was 26, 9, 5, 2.2, 2.4 and 0 per cent respectively. Only 38 (1.7 per cent) of 2184 follow-up urine cultures were positive; these results did not influence duration of follow-up. Only 26 (2.6 per cent) of 1002 contacts were infected; the yields of the first, second and third sets of cultures were 1.5, 0.6 and 0.5 per cent respectively.

Do non-steroidal anti-inflammatory drugs increase colonic permeability?

Urinary excretion of orally administered lactulose and 51 chromium labelled ethylenediamine tetra-acetate (51Cr-EDTA) was measured in 12 healthy adult subjects and in six patients with ileostomies to assess intestinal permeability. In normal subjects, 24 hour urinary recovery of 51Cr-EDTA was significantly greater than that of lactulose (mean (SEM) 2.27 (0.15) v 0.50 (0.08)% oral dose; p less than 0.001), but in ileostomy patients recovery of the two markers was the same. In normal subjects, therefore, the difference between the two markers may arise from bacterial break-down of lactulose but not of 51Cr-EDTA in the distal bowel, urinary excretion of lactulose representing small intestinal permeation and that of 51Cr-EDTA representing both small and large intestinal permeation. The markers were then given simultaneously to nine patients receiving non-steroidal anti-inflammatory drugs (NSAIDs) for rheumatoid arthritis and osteoarthritis. The 24 hour urinary recovery of 51Cr-EDTA in the patients was significantly greater than normal (4.64 (1.20) v 2.27 (0.15)% oral dose; p less than 0.01), but that of lactulose was not significantly affected. Moreover, the increase in 51Cr-EDTA recovery was most noticeable in the later urine collections. Both of these findings suggest that NSAIDs may increase colonic permeability.

Lipid peroxidation and hepatic antioxidants in alcoholic liver disease.

The generation of hepatic liver peroxidation by free radicals has been proposed as a mechanism for ethanol induced hepatotoxicity. To investigate this hypothesis, lipid extracts from hepatic needle biopsy specimens from alcoholic subjects were examined for evidence of lipid peroxidation by measuring total conjugated dienes by derivative spectroscopy and, after hydrolysis of hepatic lipid extract and reverse phase high performance liquid chromatography, the molar ratio between a diene-conjugated linoleic acid isomer (18:2 (9,11)) and the parent linoleic acid isomer (18:2(9,12)). Changes were related to hepatic histology, iron deposition, glutathione and vitamin E values. Derivative spectroscopy minima suggestive of diene conjugation were identified at 233 and 242 nm and correlated weakly, suggesting these two minima may represent different classes of lipid dienes. There was a weak relation with inflammatory histological changes in the biopsy specimen but no correlation with hepatic iron grade, glutathione, or vitamin E lipid ratio. The proportion of 18:2(9,11) linoleic acid in hepatic lipids correlated significantly with inflammatory histological features and inversely with hepatic glutathione. Furthermore, hepatic glutathione was lower in biopsy specimens with greater iron staining. The ratio of vitamin E to lipid was not related to histological group, inflammation, or iron grade. These findings suggest that excess alcohol consumption leads to hepatic inflammation and lipid peroxidation.

Measurement of conjugated diene lipids by derivative spectroscopy in heptane extracts of plasma.

A series of experiments are described which show that second derivative spectroscopy can be used to quantify conjugated lipid dienes as markers of lipid peroxidation in heptane extracts of plasma from patients with rheumatoid arthritis, osteoarthritis, and healthy controls. Results obtained by this method gave reasonable agreement with those derived from the measurement of simple absorbance in chloroform/methanol extracts. Two minima were observed in the derivative spectrum of plasma lipid extracts. These minima occurred at 233 and 241 nm and corresponded to absorbance maxima in the conventional UV spectrum. Using a combination of phospholipase hydrolysis, reverse phase high performance liquid chromatography (HPLC) and second derivative spectroscopy we confirmed that these two minima can be attributed to a single fatty acid (9 cis-, 11 trans-linoleic acid) shown previously to account for greater than 90% of diene conjugation in human plasma samples. When the biological isomer 9 cis-, 11 trans-linoleic acid was separated by reverse phase HPLC from the mixture of other plasma phospholipid-2-esterified fatty acids we observed a change in derivative spectroscopy minima from 233 and 241 nm to 228 and 237 nm. Minima at the latter two wavelengths were also seen with pure preparations of the Paint Research Isomer (9 trans-, 11 trans-linoleic acid) which eluted later than biological 9 cis-, 11 trans-linoleic acid using reverse phase HPLC, suggesting that the absorption spectra of these pure cis-, trans and trans, trans dienes are similar but can be altered by the presence of other fatty acids in the extract.

Evidence for lipid peroxidation in man following paraquat ingestion.

Four patients were investigated for evidence of lipid peroxidation between 4.5 and 36 h (mean 22 h) after ingestion of paraquat by measuring plasma phospholipid-2-esterified, diene-conjugated 18:2 delta 9,11-linoleic acid (9,11-LA) and expressing it also as a ratio R (9,11-LA/9,12-LA X 100) of the 'parent' linoleic acid. The mean value for R was 4.73 (range 3.7-7.1) at presentation and 6.91 at peak values (range 3.8-13.4) which occurred at a mean of 34 h after ingestion. Both values were significantly higher (P less than 0.001) than that of 107 healthy controls (1.94, range 0.67-3.8). Parallel changes in plasma vitamin E and 9,11-LA occurred in the 2 patients in whom serial measurements were made suggesting an involvement of vitamin E in the formation of this isomer. These findings support the hypothesis that lipid peroxidation occurs during paraquat poisoning in man and the early appearance of 9,11-LA suggests that it may be a primary event.

The use of different lipids to express serum tocopherol: lipid ratios for the measurement of vitamin E status.

Plasma tocopherol was measured in 85 alcoholic patients and 40 control subjects from a local factory. Cholesterol, triglycerides and phospholipids were measured individually and summed to give an estimate of total serum lipids. Plasma tocopherol concentrations of the alcoholics were significantly lower than those of the controls and showed wide variation from marked deficiency to the upper limit of the normal range. Using regression analysis, 1.11 mumol tocopherol/mmol total lipids were calculated as the threshold of deficiency equivalent to 0.8 mg tocopherol/g total lipid established by Horwitt et al. The sensitivity and specificity of other tocopherol:lipid ratios for identifying vitamin E deficiency was compared with the tocopherol:total lipid ratio. Thresholds of deficiency for the different tocopherol:lipid ratios were calculated. The tocopherol:cholesterol+triglyceride ratio was found to be almost as powerful in identifying vitamin E deficiency as the tocopherol:total lipid ratio (sensitivity 95%, specificity 99%). Of the tocopherol:individual lipid ratios, the tocopherol:cholesterol ratio gave the best results (sensitivity 86%, specificity 94%).