Search PubMed⌕ Search

Biomedical subjects

P Royston

Publications and source records attributed to P Royston.

At least 37 records · Page 2Linked to original sources
PubMed1999

The use of fractional polynomials to model continuous risk variables in epidemiology.

BACKGROUND: The traditional method of analysing continuous or ordinal risk factors by categorization or linear models may be improved. METHODS: We propose an approach based on transformation and fractional polynomials which yields simple regression models with interpretable curves. We suggest a way of presenting the results from such models which involves tabulating the risks estimated from the model at convenient values of the risk factor. We discuss how to incorporate several continuous risk and confounding variables within a single model. The approach is exemplified with data from the Whitehall I study of British Civil Servants. We discuss the approach in relation to categorization and non-parametric regression models. RESULTS: We show that non-linear risk models fit the data better than linear models. We discuss the difficulties introduced by categorization and the advantages of the new approach. CONCLUSIONS: Our approach based on fractional polynomials should be considered as an important alternative to the traditional approaches for the analysis of continuous variables in epidemiological studies.

Coronary Disease↗
PubMed1999

A new approach to modeling daily probabilities of conception.

Standard conception probabilities models assume that different acts of intercourse make independent contributions to the probability of conception in viable cycles. We propose an alternative, approximate model based on the assumption that the act of intercourse closest to the time of maximum fertility is the one most likely to have caused conception. We describe an adaptive algorithm [the most fertile intercourse day (MFID) algorithm] that estimates the most fertile intercourse day in each cycle. The approach is easily extended to include covariates and random between-couple differences in fecundability that affect the probability of conception in a given cycle. Reanalyses of two data sets reported in the literature are presented. Estimates of the probability of conception during the most fertile period of the cycle and of the effects of covariates are similar to estimates found using standard models.

Algorithms↗
PubMed1999

Fetal ultrasound biometry: 1. Head reference values.

OBJECTIVE: To create reliable reference ranges and calculate Z scores for fetal head ultrasound biometry using a large sample size which is evenly distributed from 12 to 42 weeks of pregnancy. DESIGN: A prospective, cross-sectional study. SETTING: Obstetric clinics (outpatient and delivery units) at the University Hospital of Zurich. SAMPLE: The study data were obtained from 6557 pregnant women. METHODS: Only the first ultrasound examination between 12 and 42 weeks of each fetus with exactly established gestational age was used for analysis. No exclusions were made on the grounds of small-for-date birthweight, prematurity or other events several weeks after the examination. Separate regression models were fitted to estimate the mean and standard deviation at each gestational age for each parameter. RESULTS: A total of 6217 fetal head biparietal diameters and 5510 occipito-frontal diameters were measured. Both head circumference and cephalic index were derived in 5462 cases where both biparietal diameter and occipito-frontal diameter could be measured on the same fetus. The centile charts, tables and regression formulae for biparietal and occipito-frontal diameters, head circumference and cephalic index are presented. An application to calculate Z scores was developed using Excel (Microsoft Corporation, USA) and macros are presented in detail in the Figure 8 footnote. The comparison of our charts with those of the two most recent studies revealed almost no differences in biparietal diameter centiles. In one publication, occipito-frontal diameter charts, and in another, head circumference charts were different from the current study. CONCLUSIONS: We have presented centile charts, tables and regression formulae for fetal head ultrasound biometry derived from a large and minimally selected sample size in a carefully designed cross-sectional study. Complete tables and regression formulae to calculate reference ranges and Z scores are presented for use in computer-aided evaluation of fetal ultrasound biometry.

Adult↗
PubMed1999

Fetal ultrasound biometry: 2. Abdomen and femur length reference values.

OBJECTIVE: To create reliable reference ranges and calculate Z scores for fetal abdomen and femur ultrasound biometry using a large sample size which is evenly distributed from 12 to 42 weeks of pregnancy. DESIGN: A prospective, cross-sectional study. SETTING: Obstetric clinics (outpatient and delivery units) at the University Hospital of Zurich. SAMPLE: The study data were obtained from 6557 pregnant women. METHODS: Only the first ultrasound examination between 12 and 42 weeks of each fetus with certainly established gestational age was used for analysis. No exclusions were made on the grounds of small-for-date birthweight, prematurity or other events several weeks after the examination. Separate regression models were fitted to estimate the mean and standard deviation at each gestational age for each parameter. RESULTS: A total of 5807 mean abdominal diameters and abdominal circumferences were derived from fetal transverse and anterio-posterior fetal abdominal diameter measurements. Fetal femur length was measured in 5860 instances. The charts, tables and regression formulae of the biometrical measurements are presented. A comparison of our charts with others showed no significant difference. Only Merz's centiles for abdominal biometry were lower and for femur length higher than ours. An application to calculate Z scores was developed using Excel (Microsoft Corporation, USA); the macros are presented in detail in the Figure 6 footnote. CONCLUSIONS: We have presented centile charts, tables and formulae for fetal abdominal diameter and circumference and femur length derived from a large and minimally selected sample size in a carefully designed cross-sectional study. Complete tables and regression formulae to calculate reference ranges and Z scores are presented to use in computer-aided evaluation of fetal ultrasound biometry.

Abdomen↗
PubMed1999

Calculating reference intervals for laboratory measurements.

Reference intervals are used in laboratory medicine to detect measurements which are extreme, possibly abnormal. Methods of estimating reference intervals and age-specific reference intervals (where the measurement is dependent on a covariate, typically age) are reviewed. The issues of calculating confidence bands, determining appropriate sample sizes and assessing goodness-of-fit are discussed.

Adult↗
PubMed1998

How to construct 'normal ranges' for fetal variables.

A method to estimate reference intervals which depend on gestational age is described. The approach is based on standard statistical techniques, so implementation should be straightforward with most statistical packages. The reference limits are defined by simple formulae which may be plotted as smooth curves. The method is illustrated with analyses of three real datasets arising from different aspects of fetal monitoring.

Cephalometry↗
PubMed1997

Simplified estimation of age-specific reference intervals for skewed data.

Age-specific reference intervals are commonly used in medical screening and clinical practice, where interest lies in the detection of extreme values. Many different statistical approaches have been published on this topic. The advantages of a parametric method are that they necessarily produce smooth centile curves, the entire density is estimated and an explicit formula is available for the centiles. The method proposed here is a simplified version of a recent approach proposed by Royston and Wright. Basic transformations of the data and multiple regression techniques are combined to model the mean, standard deviation and skewness. Using these simple tools, which are implemented in almost all statistical computer packages, age-specific reference intervals may be obtained. The scope of the method is illustrated by fitting models to several real data sets and assessing each model using goodness-of-fit techniques.

Adolescent↗
PubMed1997

Volumetric self-sampling of cervicovaginal fluid to determine potential fertility: a multicentre pre-effectiveness study of the Rovumeter.

The aim of this study was to assess how effectively the Rovumeter, designed for the volumetric self-sampling of cervicovaginal fluid (CVF), can be used to locate the minimum period of potential fertility (PPF) during ovulatory cycles. A multicentre, prospective study was undertaken of volunteers (attending natural family planning clinics) over three consecutive, apparently normal, menstrual cycles. All women collected daily samples of early morning urine and CVF and recorded the volumes (to the nearest 1.0 and 0.1 ml respectively). The concentrations of oestrone glucuronide (EG), luteinizing hormone (LH) and pregnanediol glucuronide (PG) were measured in all samples of early morning urine by immunoassay. A preliminary data set was used to optimize an algorithm to detect the start and end of potential fertility from the volumes of CVF. The end-points used were the normality of each menstrual cycle from its length, the length of luteal phase, and concentrations of EG, LH and PG, the start and end days of potential fertility from CVF volumes, and the minimum PPF, which was defined as the day of the LH peak minus 3 to day plus 2 inclusive. Overall, 72 women (median age 30 years, range 24-38) were recruited from three centres (23 from Birmingham, 24 from Milan, 25 from Santiago) and contributed data from 235 menstrual cycles (median length 28 days, range 23-44). The urinary LH peak was identified in 228 cycles (97%; median time, day 15 from day 1 of last menses, with range day 10 to day 35). The use of the Rovumeter gave start and end signals of potential fertility during 138 cycles (59%). The median length of the derived PPF was 8 days (range 4-18). The signals covered the defined, minimum PPF in 113 cycles [i.e. 50% of those with an LH peak; range 28% (Milan) to 62% (Birmingham)]. Overall 16/72 women (22%) had successful tests over three consecutive menstrual cycles [range 2/24 (8%; Milan) to 8/23 (35%; Birmingham)]. We conclude that signals from daily changes in the volume of CVF as determined by the use of the Rovumeter consistently locate the minimum period of potential fertility in only a small proportion of women.

Adult↗
PubMed1996

A method for determination in situ of variations within the hepatic lobule of hepatocyte function and metabolite concentrations.

A method is described for the production of detailed maps of intralobular variations of hepatocyte function and metabolite concentrations, based on variable destruction by digitonin of the lobule from the centrilobular direction. Instead of the conventional approach, in which isolated hepatocytes are then prepared and studied in suspension, perfusion is continued after digitonin treatment and the function of the unaffected lobular remnants is determined, or mean metabolite concentrations are measured by 31P-NMR. These measurements are plotted against the degree of destruction, determined precisely after each study by automated quantitative histomorphometry. These plots are transformed into curves of the function or metabolite concentration of nominal single cells at any point along the radius of the lobule. Gluconeogenesis from lactate remained stable, although reduced, even after 85-90% lobular destruction, predominated periportally and disappeared by 50% along the radius of the lobule. In 31P-NMR studies, employing 1.5 mM lactate as substrate, narrowing of the intracellular P1 resonance was observed as digitonin destruction increased; this was attributed to a decrease in the intralobular heterogeneity of the intracellular pH, which fell from approx. 7.9 to < 7.4 along the first 16% of the lobular radius (from the periportal end) and to < 7.3 in the remainder of the lobule. The ATP concentration rose, and then fell, along the radius of the lobule in a centripetal direction. The method is potentially generally applicable to a wide range of hepatocellular functions and to the measurement of metabolite concentrations, most conveniently those susceptible to estimation by NMR.

Animals↗
PubMed1996

Comparison of 15 mg and 25 mg of bupivacaine both with 50 microg fentanyl as initial dose for epidural analgesia.

Bupivacaine 15 mg is commonly used as a test dose prior to the initial dose for epidural analgesia in labour. When 15 mg or 25 mg of bupivacaine (15 ml of 0.1% or 0.167%) with 50 microg fentanyl was administered blindly to two groups of labouring women, as an initial dose, 83% and 90% of women respectively, achieved analgesia within 20 min. All the remaining women achieved analgesia with a further dose of 10 mg bupivacaine (10 m10.1%) with fentanyl 2 microg per ml given at 20 min.

Journal Article↗
PubMed1995

Calculation of unconditional and conditional reference intervals for foetal size and growth from longitudinal measurements.

Methods for estimating unconditional and conditional reference intervals for foetal size and growth based on longitudinal observations are presented. The methods are based on simple random-effects regression models and involve transforming both the response and the covariate (timepoint). A dataset from a designed longitudinal study of foetal size is analysed in detail as a motivating example.

Anthropometry↗
PubMed1995

Epidermal growth factor is digested to smaller, less active forms in acidic gastric juice.

BACKGROUND/AIMS: Epidermal growth factor (EGF) is present in gastric juice and has potent mitogenic properties. The stability of EGF in gastric juice under various physiological and pathophysiological conditions was examined. METHODS: Recombinant human EGF1-53 was incubated with HCl containing pepsin. We also determined the forms of EGF present in the gastric juice of patients under basal conditions, patients taking the acid suppressant omeprazole, patients with achlorhydria, and volunteers undergoing intragastric neutralization with NaHCO3 (n = 6 per group). Samples were analyzed using mass spectroscopy and/or high-pressure liquid chromatography followed by radioimmunoassay. The effect of acid and pepsin digestion on EGF bioactivity was determined using an in vitro hepatocyte bioassay and an in vivo cytoprotection assay in the rat stomach. RESULTS: EGF1-53 was digested to the EGF1-49 and EGF1-46 forms in all samples containing pepsin when the pH was < 4. In gastric juice samples with pH > 4, the proportion of intact EGF increased to about 60%. For both methods of bioassay, intact EGF1-53 was about 3-4 times as potent as acid and pepsin-treated EGF. CONCLUSIONS: EGF is produced in the 1-53 form but is rapidly cleaved to smaller, less active forms in acidic gastric juice. In contrast, only a small proportion of the EGF is cleaved if the pH is maintained above 4. This mechanism may be relevant to the healing process of acid suppressants.

Achlorhydria↗
PubMed1995

Comparing non-nested regression models.

A method for comparing the fits of two non-nested models, based on a suggestion of Davidson and MacKinnon (1981), is developed in the context of linear and nonlinear regression with normal errors. Each model is regarded as a special case of an artificial "supermodel" and is obtained by restricting the value of a mixing parameter gamma to 0 or 1. To enable estimation and hypothesis testing for gamma, an approximate supermodel is used in which the fitted values from the individual models appear in place of the original parametrization. In the case of nested linear models, the proposed test essentially reproduces the standard F test. The calculations required are for the most part straight-forward (basically, linear regression through the origin). The test is extended to cover situations in which serious bias in the maximum likelihood estimate of gamma occurs, simple approximate bounds for the bias being given. Two real datasets are used illustratively throughout.

Biometry↗
PubMed1993

Screening for early familial ovarian cancer with transvaginal ultrasonography and colour blood flow imaging.

OBJECTIVE: To assess the value of transvaginal ultrasonography with colour blood flow imaging in detecting early ovarian cancer in women with a family history of the disease. DESIGN: Study of self referred symptomless women with a close relative who had developed the disease. Each woman was screened to detect persistent lesions and defined changes in ovarian volume. Morphological score and pulsatility index were recorded. SETTING: Ovarian screening clinic. SUBJECTS: 1601 self referred women. INTERVENTIONS: Women with a positive screening result were recommended to have further investigations. MAIN OUTCOME MEASURES: Findings at surgery and histology of abnormal ovaries. Morphological score > or = 5 and pulsatility index < 1.0 at last scan. RESULTS: Women were aged 17 to 79 (mean 47) years; 959 (60%) were premenopausal, 469 (29%) were naturally postmenopausal, and 173 (11%) had had a hysterectomy. 157 women had a pedigree suggestive of the site specific ovarian cancer syndrome and 288 of multiple site cancers. 61 women had a positive screening result (3.8%, 95% confidence interval 2.9 to 4.9%), six of whom had primary ovarian cancer detected at surgery (five stage Ia, one stage III). Use of a high morphological score or a low pulsatility index increased the odds of finding ovarian cancer from 1:9 to about 2:5 (1:1 in the highest risk groups). Five interval cancers were reported (three ovarian and two peritoneal). Eight of the 11 cancers developed in women with pedigrees suggestive of inherited cancer. CONCLUSIONS: Transvaginal ultrasonography with colour flow imaging can effectively detect early ovarian cancer in women with a family history of the disease. The screening interval should be less than two years.

Adolescent↗
PubMed1993

Exact conditional and unconditional sample size for pair-matched studies with binary outcome: a practical guide.

Tables of sample sizes for pair-matched studies with binary outcome are presented. They are based on conditional and unconditional approaches using the 'exact' (binomial) test. An approximate procedure is suggested to estimate the sample size for parameter values that do not correspond exactly with table entries. The procedure utilizes a minor modification of the large-sample formula given by Connett et al. A practical strategy for estimating the overall sample size in the presence of a nuisance parameter (the proportion of discordant pairs) is recommended. An example from a proposed clinical trial is given.

Binomial Distribution↗