Clive J. Lawrence

Plus ça change: predictors of birthweight in two national studies

Summary. The 16 989 singleton births in one week of March, 1958, studied by the British Perinatal Mortality Survey, were subjected to an analysis of covariance, which showed that major factors associated with birthweight of the infant were: maternal height, history of smoking in pregnancy, parity and history of pre‐eclampsia during the pregnancy. The same analysis was repeated on the data collected on 16 792 singletons born 12 years later in one week of April, 1970 and studied by the British Births Survey. In spite of major changes in obstetric practice and in the maternal population, the same factors were shown to be highly significant and the magnitude of the associations had changed little.

Are postbereavement research interviews distressing to carers? Lessons learned from palliative care research

Research into the quality and impact of palliative care for patients and carers has expanded dramatically over the past 25 years. The work of Parkes was prominent in the use of postbereavement interviews in the assessment of the impact of the death and dying of a patient on the surviving spouse.1,2 A number of large, methodologically rigorous studies of caring for the dying have been published which involved interviewing bereaved carers.3–5 Such interviews can be viewed as being either intrusive or beneficial to the carer and this has implications when seeking ethical approval for such studies.5 This short report seeks to show some of the lessons learned from a research project designed to examine terminal cancer care provided by community hospitals and an inpatient hospice.

Characteristics of referrals to an inpatient hospice and a survey of general practitioner perceptions of palliative care.

In order to determine symptoms, drug prescribing and physical problems of patients referred to an inpatient hospice, case notes from 130 consecutive first admissions (95 general practitioner (GP) referrals, 35 consultant referrals) were analysed. GP referrals were more likely to be constipated, require care and be discharged to home. Consultant referrals were more gravely ill, dependent and more likely to die in the hospice. On admission 76 (58%) patients were receiving opiates with co-prescription of opiate and laxative occurring in 41 % (31/76) of the cases. The prescription of laxatives with the symptoms of constipation occurred in 62% (26/42) of the cases on admission. A telephone survey of 79 referring GPs revealed that 37% had attended neither a course nor a lecture relevant to palliative care in the past 3 years. GPs experienced difficulties frequently or always in: (a) managing pain (8/79); (b) managing other symptoms (25/79); (c) helping patients and relatives cope with their emotional distress (18/79); and (d) coping with their own emotional responses to death and dying (5/79). In conclusion, the differences demonstrated between the GP and consultant referrals have implications for purchasers. The high incidence of possible opiate-induced side-effects and the difficulties with symptom control expressed by some GPs indicate a continuing need for effective educational input.

The influence of general practitioner community hospitals on the place of death of cancer patients

All deaths from cancer were identified from death certificates in the Exeter Health District for a period of one year. Place of death, age, cancer type and access to general practitioner community hospital beds and the domiciliary hospice service were recorded. There were 1022 deaths attributable to cancer (parts 1 a, 1 b or 1 c of the death certificate) who were patients of general practitioners in the health district. The place of death for patients with access to community hospital beds were: home 173/590(29%), community hospital 232/590 (39%), specialist services unit 102/590 (17%), nursing or residential home 32/590 (5%), Marie Curie hospice 51/590 (9%). For patients without access to community hospital beds the place of death was: home 177/427 (41 %), specialist services unit 165/427 (39%), nursing or residential home 42/427 (10%), Marie Curie hospice 43/427 (10%). The presence of community hospital beds was associated with a significant reduction of deaths in the specialist service unit (p<0.001) and with a smaller reduction in home deaths (p<0.01). Access to the domiciliary hospice services in areas with community beds was not associated with any significant change in the place of death. General practitioners cared for 74% of cases at the time of death in areas with access to community hospital beds and for 51 % of cases without such access, which was a significant difference (p <0.001). It therefore appears that community hospitals play a major role in the terminal care of cancer patients and access to such beds is associated with a decrease in cancer deaths occurring in specialist services beds.

Delayed onset of regular respiration and subsequent development.

A comprehensive survey carried out at birth, data on mortality and a 5 year follow-up covering medical, educational and sociological aspects of child development were available for singleton births born in one week of April 1970. The survey at 5 years of age included 12363 children, 79.6% of the surviving cohort. An assessment has been made of the extent to which delay in the onset of regular respiration at birth is associated with the subsequent development of the child. Using the results of our previous investigations, the relationships between this delay and other factors known at the time of birth have been taken into account. The distributions of a large number of development variables were not significantly different for groups of children determined by respiratory delay at birth, but there was an association with mortality and cerebral palsy. In addition, there was an unexpected relationship between delayed onset of respiration and subsequent bronchitis. This association remained significant after controlling for possible confounding effects using linear modelling techniques.

Sodium in hand and pump expressed human breast milk.

Sodium content was analysed in the expressed breast milk from 30 mothers. Measurements were taken up to 24 days postpartum, during which time the mothers expressed manually or by means of a pump. Statistical analysis has confirmed not only the known decline in sodium levels postpartum but has also shown a significantly higher sodium content in manually expressed milk compared to that from pump expressed milk. This finding suggests that the method of expression may alter the sodium concentration in human breast milk, thus potentially reducing the amount of sodium supplementation required by some preterm infants.

Factors associated with delayed onset of regular respiration

In a population of 16333 singleton births born in one week of April, 1970, 4.4% failed to establish regular respiration within 3 min of birth. A follow-up at 5 years of age collected medical, educational and sociological information on 79.6% of the survivors of the cohort identified by the birth survey. Data have been analysed to investigate the relationships between the delay in the onset of regular respiration and possible risk factors identifiable prior to or at the time of birth. Complex statistical analysis employing linear modelling techniques showed that the delay in the onset of respiration was primarily related to: null and high parity; antepartum haemorrhage (especially placenta praevia); pre-eclampsia; breech delivery; Caesarian section and low birthweight. There was no association with maternal smoking during pregnancy or with social class. These conclusions are reasonably consistent with those drawn from other studies using alternative measures of asphyxia.

Comparison of traditional GP oral anticoagulation management with a nurse-led service involving near patient testing and computerized decision support

The use of oral anticoagulation therapy in primary care is increasing. We compared general practitioner (GP)-led oral anticoagulation monitoring with a nurse-led service involving near patient testing and computerized decision support (NPT-CDS). The nurse-led NPT-CDS service provided anticoagulation control equivalent to the traditional GP-led service, with identical mean international normalized ratio (INR) values and a nonsignificant trend towards improvement in all other parameters. Recording of indications for anticoagulation and target INR ranges were significantly improved using CDS software. For patients established on warfarin, the GP-led service was costed at £56.88 per patient per year, compared with £63.76 for the nurse-led NPT-CDS service. Patients overwhelmingly preferred the NPT-CDS service for reasons of convenience, avoidance of phlebotomy and improved dosage instructions. Performance of the NPT-CDS service within the National External Quality Assurance Scheme (NEQAS) was satisfactory. In conclusion, nurse-led oral anticoagulation utilizing NPT-CDS is an effective and acceptable alternative to traditional GP-led monitoring. The costs of an NPT-CDS service are higher, but need to be set against factors such as patient satisfaction and escalating GP workload.

Noise and White Noise

Noise and white noise arise in the description of time series data. Both are defined in this entry. Keywords: interference; stationary; communication system

Robust Methods in Time Series Analysis

In this article, a number of techniques for analyzing time series data that are robust to the presence of outliers are discussed. The general approach is to remove or suppress the possible effects of contamination or spurious information by smoothing. Keywords: time series; smoothing; neural networks; robust Bayesian estimation