Library

Notes: Objective By combining serial measurements of the circulating concentrations of thromboxane A, and prostacyclin with measurements of venous distensibility (taken during the pregnancies of both normal women and those with pregnancy induced hypertension or pre-eclampsia), to test the following hypotheses: 1. that changes in the venous plasma ratio of thromboxane (TXB2) and 6-keto-PGF1α would correlate with changes in the blood pressure of women developing and recovering from pregnancy induced hypertension or pre-eclampsia and 2. that changes in venous distensibility would correlate with changes in arterial blood pressure in pregnancy induced hypertension or pre-eclampsia.Design Prospective, longitudinal cohort study.Setting John Hunter Hospital clinic, Newcastle, Australia.Subjects One hundred and sixty primiparous women, recruited when presenting for their first routine antenatal visit, were investigated at, or close to, 19, 28 and 37 weeks of gestation; a subgroup was also studied in the postnatal period. The measurements of the patients who developed pregnancy induced hypertension or pre-eclampsia were compared with those of controls selected from the cohort.Main outcome measures Serial measurements of the circulating concentrations of the stable metabolites of thromboxane A2 and prostacyclin (TXB2 and 6-keto-PGFlα, respectively), venous distensibility and immediate (no rest) and resting (for at least 30 min) blood pressures.Results There was no significant difference between the subject and control groups at any time during or after the pregnancy in the concentrations of prostaglandin metabolites, their ratio or venous distensibility. In contrast, there was a significant difference between the groups at 19 weeks for immediate and resting readings of diastolic pressure (6 mmHg (95 % CI 1.5 to 10.5) and 4 mmHg (95% CI 0.1 to 7.9), respectively). These differences increased through the pregnancy but mean postnatal readings for the groups were almost identical suggesting that the subjects were not intrinsically hypertensive compared with controls. Blood pressures for the subject group, both immediate and resting, were significantly different from the 19 week readings at 28 weeks (diastolic) and at 37 weeks (systolic and diastolic). The only significant change from first readings among controls was in postnatal systolic pressure which was significantly higher than 19 week values, probably reflecting the vasodilatation, with accompanying hypotension, of early, normal pregnancy. This difference was not observed in those who subsequently developed pregnancy induced hypertension or pre-eclampsia.Conclusions Our study was unable to demonstrate differences in circulating metabolites or venous distensibility between normotensive women and those with pregnancy induced hypertension or pre-eclampsia. If pregnancy induced hypertension or pre-eclampsia in humans represents not so much the presence of abnormal constrictor influences as a process initiated by failure of normal vasodilatation in early pregnancy, studies carried out later may detect mainly adaptive and secondary changes.

Notes: 1. There is little hypothesis-testing clinical research performed in toxicology. Randomized clinical trials are rare and most observational studies are performed on highly selected patients and are subject to marked bias. Thus, for many poisonings, our approach has been based almost entirely on deduction from known pharmacological/toxicological effects, generalizations from drugs within the same therapeutic class, animal data and case reports. This is also far from satisfactory, as many toxicological mechanisms are poorly understood and not related to the therapeutic class.2. Although we need much better data to address the clinical and public health aspects of poisoning, there are many practical and ethical reasons why randomized clinical trials are difficult in this field. However, the scope for observational research, in particular population-based clinical epidemiology, is almost unlimited. The collection of data on human poisoning is facilitated because most non-fatal overdoses are admitted to hospital and by legal requirements to report to the coroner deaths that are due to poisoning. In the present article I argue that ‘toxicoepidemiology’, meaning the application of epidemiological methods to the problem of acute poisoning, is the best means we have of addressing deficiencies in our knowledge of poisoning.3. Examples are given of a variety of observational research strategies, ranging from audit to meta-analysis, that may be applied to clinical toxicology. From coronial and clinical data obtained from reasonably well-defined populations, it has been possible to identify a number of previously unrecognized differences in the severity and spectrum of toxicity between and within drug classes. Also, the demographic risk factors for poisoning and the reproducibility, validity and optimal use of diagnostic and therapeutic interventions can be assessed.4. The major limitations to the range of associations and interventions that may be studied are the need to achieve adequate power to study uncommon outcomes or poisonings and the ability to replicate findings at other centres using similar methodology. The expansion of data collection to other centres has the potential largely to overcome these obstacles.