Elisabeth Svensson

Guidelines to statistical evaluation of data from rating scales and questionnaires.

Questionnaires and rating scales are commonly used to measure qualitative variables, such as feelings, attitudes and many other behavioural and health-related variables. There are different types of instruments ranging from single scales to multidimensional, multi-item questionnaires. The scaling of the responses can vary from the dichotomous alternatives “yes” and “no” to a mark on a line, as in the visual analogue scale (VAS). Numerical labels are commonly used for the recordings. Nevertheless, irrespective of the type of scaling, the item responses indicate only an ordered structure and not a numerical value in a mathematical sense. Such data are often called ordered categorical or ordinal (1–4). Statistical methods for data from rating scales must take account of the rank-invariant propertiesof ordinal data, whichmeans that the methodsmust be unaffected by a relabelling of the scale categories. Hence, statistical methods applicable to data from rating scales differ completely from the traditional methods for quantitative variables, since calculations based on adding or subtracting ordinal data are not appropriate. Sum scores of multi-item assessments, the mean value, standard deviation and calculation of differences for description of change in score do not have an interpretable meaning and must be avoided in the statistical evaluation of data from rating scales and questionnaires (4–6). Traditionally, in applied research, there is a temptation to treat data from rating scales as numerical on an interval level (4, 5). It should be emphasized, however, that data on an interval level are quantitative, which means that such data have the mathematical properties of well-dex8e ned size and equidistance, but the same variable does not have the same ratio when it is measured in different units (1). Hence, qualitative data could never gain the properties required for being treated as interval data. Statistical methods for quantitative data are valid only when data have the mathematical properties of well-dex8e ned size and distance, and conclusions drawn from such analyses are solely interpretable and reliable for quantitative data. However, quantitative data such as blood pressure could be treated as ordinal when categorized as “low”, “normal” and “high”. Such categorization changes the choice of appropriate statistical methods of analysis. GUIDELINES FOR STUDIES INCLUDING RATING SCALES AND/OR QUESTIONNAIRES

Moderate alcohol intake increases fibrosis progression in untreated patients with hepatitis C virus infection

Although excessive alcohol consumption in combination with hepatitis C virus (HCV) infection is known to increase the risk of liver cirrhosis, the effect of moderate alcohol intake remains to be elucidated. The aim of this study was to evaluate the effect of moderate alcohol consumption on fibrosis progression in HCV infection. A group of 78 patients with HCV infection and moderate alcohol consumption were analysed retrospectively. All patients had undergone two liver biopsies, with a median time between biopsies of 6.3u2003years, and had not received any antiviral therapy. Their lifetime drinking history was recorded. All patients except one had daily alcohol consumption below 40 g of ethanol (median 4.8u2003g/day, interquartile range 1.1–11.6u2003g/day) during the period between the biopsies. The patients whose liver fibrosis had deteriorated had a higher total alcohol consumption and higher drinking frequency between the biopsies. The degree of fibrosis progression was greater in patients with a total alcohol intake and drinking frequency above the median level for the group. A multiple logistic regression analysis showed that drinking frequency and time between biopsies were independently associated with fibrosis progression. Hence, even moderate alcohol intake seems to increase fibrosis progression in HCV‐infected patients. From that point of view, total abstention ought to be recommended. If this is not achieved, occasional use of alcohol is probably less harmful than daily drinking for patients with low or moderate alcohol consumption.

Low Gestational Age Associated with Abnormal Retinal Vascularization and Increased Blood Pressure in Adult Women

The objective was to investigate any possible relationship between functional and structural vascular changes in women with low gestational age and/or low birth weight by analyzing the retinal vascular pattern in women with thoroughly documented blood pressure. Retinal vessel morphology was evaluated by digital image analysis of ocular fundus photographs in 47 subjects, aged 23–30 y. The women were allocated into three groups: 1) those born preterm and appropriate for gestational age (AGA), with a median gestational age at birth of 30 wk and a median birth weight of 1250 g (n = 14); 2) those born small for gestational age (SGA) but full term (median 40 wk), with a median birth weight of 2130 g (n = 17), and 3) those born full term, AGA, and with a median birth weight of 3640 g (n = 16). Women born preterm had significantly higher length index for arterioles compared with the other two groups (median 1.11 and 1.08, respectively, p = 0.005). In addition, the preterm-born women had significantly fewer number of vascular branching points compared with the controls (median 27 and 30, respectively, p = 0.03). The abnormal retinal vascularization observed in ex-preterm women together with an increased casual blood pressure observed in these subjects suggests that being born preterm does have effects on the vascular system that persist into adult life. In addition, it demonstrates that preterm birth seems to affect the vascular system both functionally and structurally, which, in adulthood, could result in a lower threshold for the development of vascular disease.

Comparison of the Quality of Assessments Using Continuous and Discrete Ordinal Rating Scales

There are various types of rating scales in which the possible responses are either continuous or discrete. A large number of studies have compared the visual analogue scale (VAS) with various types of discrete scales, such as verbal descriptor scales and numerical rating scales. These studies have reported conflicting results, with no firm evidence of superiority of either type of scale. However, the conclusions have often been based on parametric statistical methods. In the present study, a rank-invariant approach for inter-scale comparisons is used to evaluate the order consistency between the VAS, the graphic rating scale and a five-point verbal descriptor scale. The comparative evaluation showed that both the verbal descriptor scale and the graphic rating scale assessments were superior to the visual analogue scale assessments, and that assessments on the discrete scale had the highest level of stability.

Concordance between ratings using different scales for the same variable

Qualitative variables are commonly measured by means of different types of rating scales that produce ordered categorical data. The number of response categories in a scale can be chosen arbitrarily according to different operational definitions of the attribute to be measured. In the present paper classical measures based on the difference between the probabilities of concordant and discordant pairs of classifications were compared with a novel approach for assessment of order consistency between rating scales. In the new approach, illustrated by an assessment of pain, the concordant and discordant pairs are related to the pattern of total order consistency, irrespective of the scaling and the categorical distributions. The level of order consistency between a visual analogue scale (VAS) and a discrete scale for the same variable was evaluated. The non-linear properties of VAS assessments were demonstrated. The inter-scale consistency between the discrete scale and different approaches to grouping VAS responses into discrete scales showed that equidistant rescaling of VAS assessments resulted in an inter-scale bias.

The clinical and morphologic spectrum of optic nerve hypoplasia

PURPOSEnThe purpose of this study was to characterize the clinical and morphologic spectrum of all children referred for optic nerve hypoplasia to a tertiary referral hospital in Sweden during a 9-year period.nnnSUBJECTS AND METHODSnA retrospective review was undertaken of the charts of 117 children (age range, 0.25-16 years), treated at the Childrens Hospital, Göteberg between 1988 and 1996, after the diagnosis of optic nerve hypoplasia. Ocular fundus morphologic condition was evaluated by digital image analysis of fundus photographs in 50 children, and neuroimaging was performed in 57 children.nnnRESULTSnOf the 117 children with optic nerve hypoplasia, 66 (56%) were boys and 51 (44%) were girls. Preterm birth occurred in 24 (20%), and 14 (12%) were born small for gestational age. Additional diagnoses, such as fetal alcohol syndrome, septo-optic dysplasia, perinatal adverse events, and neuropsychiatric disorders, were made in 88%; 7% had unilateral optic nerve hypoplasia. Most of the children had small optic disc, cup, and neuroretinal rim areas, as well as retinal vascular abnormalities; 75% were visually impaired, and a high incidence of nystagmus and strabismus was found among these children.nnnCONCLUSIONnThis study indicates that optic nerve hypoplasia has a wide clinical and morphologic spectrum and is associated with a broad range of disorders of the central nervous system. It is suggested that differences in the etiology and timing of the lesion as well as associated lesions may explain this spectrum of optic nerve hypoplasia in children.

Ocular fundus abnormalities in children born before 29 weeks of gestation: A population-based study

Purpose: Preterm birth has been found to be associated with increased morbidity of the central nervous and vascular tissues. To investigate the influence of preterm birth on the optic disc and retinal vessels, we examined the ocular fundus in school-aged children born very preterm.Methods: A prospective, population-based study was performed in 50 very preterm children (median age 7 years, range 5–9 years) with a median gestational age at birth of 27 weeks (range 24–28 weeks) and a median birth weight of 1055 g (range 450–1520 g). The ocular fundus was examined by ophthalmoscopy in 50 children, and the optic nerve and retinal vessel morphology was evaluated by digital image analysis of ocular fundus photographs in 45 of these children.Results: The median optic disc area was significantly smaller (p = 0.0002) in the preterm children compared with a reference group. There was no difference in cup area and, consequently, the rim area was significantly smaller (p = 0.0002) in the preterm children. Children with early signs of brain lesions commonly had a rim area below the median of the reference group. Preterm children also commonly had an abnormal retinal vascular pattern that was independent of a previous history of retinopathy of prematurity.Conclusion: Very preterm birth was associated with subnormal optic disc and rim areas and an abnormal vascular pattern. The findings clearly demonstrate the effect of preterm birth on the development of these structures. The long term clinical prognosis of these findings has yet to be determined.

Optic disc morphology may reveal timing of insult in children with periventricular leucomalacia and/or periventricular haemorrhage

Aims: To evaluate the relation between optic disc morphology and timing of periventricular white matter damage, defined as either periventricular leucomalacia (PVL) or periventricular haemorrhage (PVH), as estimated by neuroradiology. Methods: 35 children with periventricular white matter damage who had had neuroradiology performed and ocular fundus photographs taken had their photographs analysed by digital image analysis and compared with a control group of 100 healthy full term children. Timing of brain lesion was estimated by analysis of the brain lesion pattern on neuroradiological examinations (magnetic resonance imaging or computed tomography). Results: Four of 35 children had a small optic disc area; these four children had a brain lesion estimated to have occurred before 28 weeks of gestation. Nine of 11 children with a large cup area had a PVL/PVH estimated to have occurred after 28 weeks of gestation. The children with PVL/PVH had a significantly larger cup area (median 0.75 mm2) than the control group (median 0.33 mm2) (pu200a=u200a0.001) and a significantly smaller neuroretinal rim area (median 1.58 mm2) than the controls (median 2.07 mm2) (pu200a=u200a0.001). Conclusion: In a child with PVL/PVH and abnormal optic disc morphology, the possibilities of timing of the lesion should be considered.

Analysis of interobserver disagreement in the assessment of subarachnoid blood and acute hydrocephalus on CT scans

The purpose of this study was to analyse factors for interobserver disagreements in two scales used for the assessments of the amount of blood in subarachnoid space (Fisher grading) and of acute hydrocephalus on computerized tomographic (CT) scans. The assessments made by four neuroradiologists on 59 CT scans obtained in the acute stage after subarachnoid hemorrhage were analysed by a statistical method by Svensson and Holm. This method permits the separation of the inter-observer disagreements in their random and systematic components. The overall consistency of the assessments was significant (p < 0.0005) but the neuroradiologists disagreed on half of the CT-scans. The kappa values were 0.50-0.63. The analysis showed that the main reason for disagreements was systematic inter-observer differences in their use of the clinically most important parts of staging, i.e. subarachnoid clot or intraventricular blood (Fisher grading) and too low categories (hydrocephalus). The main conclusion from this study is that the proper remedy for Fisher grading and for grading of hydrocephalus is a sharpening of the criteria of specific category levels and given this improvement both grading systems will show a high level of reliability.

Diagnostic value of magnetic resonance imaging and planimetric measurement of optic disc size in confirming optic nerve hypoplasia.

PURPOSEnThis study set out to evaluate magnetic resonance imaging (MRI) and measurement of optic disc size as diagnostic tools for confirming the diagnosis of optic nerve hypoplasia (ONH) in children with impaired growth.nnnMETHODSnMRI was performed to study the size of the intracranial visual pathways, and image analysis measurements of fundus photographs were performed to determine the size of the optic disc. Results from these investigations were compared with those using the gold standard for diagnosis of ONH, which was the clinical eye examination (visual function, ophthalmoscopic signs, or both). Forty children (median age, 9 years; range, 3 to 19 years) with impaired growth were included in the study.nnnRESULTSnThe prevalence of ONH among the children was 15%. MRI classification of the visual pathways had a higher positive predictive value than image analysis measurement of the optic disc size (1.0 vs 0.6).nnnCONCLUSIONSnMRI is a good tool for confirming the diagnosis of ONH and may thus facilitate early detection. On the other hand, a small optic disc per se is not a definite indicator of ONH but should encourage further investigation with MRI, especially if there is a clinical suspicion of ONH.