Knowledge may not be the best target for strategies to influence evidence-based practice: using psychological models to understand RCT effects.

BACKGROUND: Interventions to enhance the implementation of evidence-based practice have a varied success rate. This may be due to a lack of understanding of the mechanism by which interventions achieve results. PURPOSE: Use psychological models to further an understanding of trial effects by piggy-backing on a randomised controlled trial testing 2 interventions (Audit & Feedback and Computer-aided Learning) in relation to evidence-based third molar management. METHOD: All participants of the parent trial (64 General Dental Practitioners across Scotland), regardless of intervention group, were invited to complete a questionnaire assessing knowledge and predictive measures from Theory of Planned Behaviour and Social Cognitive Theory. The main outcome was evidence-based extracting behaviour derived from patient records. RESULTS: Neither intervention significantly influenced behaviour in the parent trial. This study revealed that the interventions did enhance knowledge, but knowledge did not predict extraction behaviour. However, the interventions did not influence variables that did predict extraction behaviour (attitude, perceived behavioural control, self-efficacy). Results suggest both interventions failed because neither influenced possible mediating beliefs for the target behavior. CONCLUSION: Using psychology models elucidated intervention effects and allowed the identification of factors associated with evidence based practice, providing the basis for improving future intervention design.

Effectiveness of strategies to disseminate and implement clinical guidelines for the management of impacted and unerupted third molars in primary dental care, a cluster randomised controlled trial.

OBJECTIVE: To investigate the effectiveness and cost-effectiveness of different guideline implementation strategies, using the Scottish Intercollegiate Guidelines Network (SIGN) Guideline 42 "Management of unerupted and impacted third molar teeth" (published 2000) as a model. DESIGN: A pragmatic, cluster RCT (2x2 factorial design). SUBJECTS: Sixty-three dental practices across Scotland. Clinical records of all 16-24-year-old patients over two, four-month periods in 1999 (pre-intervention) and 2000 (post-intervention) were searched by a clinical researcher blind to the intervention group. Data were also gathered on the costs of the interventions. INTERVENTIONS: Group 1 received a copy of SIGN 42 Guideline and had an opportunity to attend a postgraduate education course (PGEC). In addition to this, group 2 received audit and feedback (A and F). Group 3 received a computer aided learning (CAL) package. Group 4 received A and F and CAL. PRINCIPAL OUTCOME MEASUREMENT: The proportion of patients whose treatment complied with the guideline. RESULTS: The weighted t-test for A and F versus no A and F (P=0.62) and CAL versus no CAL (P=0.76) were not statistically significant. Given the effectiveness results (no difference) the cost effectiveness calculation became a cost-minimisation calculation. The minimum cost intervention in the trial consisted of providing general dental practitioners (GDPs) with guidelines and the option of attending PGEC courses. Routine data which subsequently became available showed a Scotland-wide fall in extractions prior to data collection. CONCLUSION: In an environment in which pre-intervention compliance was unexpectedly high, neither CAL nor A and F increased the dentists' compliance with the SIGN guideline compared with mailing of the guideline and the opportunity to attend a postgraduate course. The cost of the CAL arm of the trial was greater than the A and F arm. Further work is required to understand dental professionals' behaviour in response to guideline implementation strategies.

Can psychological models bridge the gap between clinical guidelines and clinicians' behaviour? A randomised controlled trial of an intervention to influence dentists' intention to implement evidence-based practice.

OBJECTIVE: The lag between publication of evidence for clinical practice and implementation by clinicians may be decades. Research using psychological models demonstrates that changing intention is very important in changing behaviour. This study examined an intervention (rehearsing alternative actions) to change dentists' intention to implement evidence-based practice (EBP) for third molar (TM) management. DESIGN: Randomised controlled trial / postal. SETTING: Primary care. SUBJECTS AND METHODS: Dentists were randomly selected from the Scottish Dental Practice Board Register, then randomly allocated to intervention or control groups, and sent a questionnaire. The intervention group listed management alternatives to TM extraction prior to their TM extraction intention, and the control group did not. Based on psychological models for reducing a behaviour's frequency (EBP is weighted against TM extraction), prior listing of alternatives should decrease extraction intention. MAIN OUTCOME MEASURES: Intention to extract TMs. RESULTS: A total of 99 dentists - 70 Males, 29 Females; mean age = 41.42 years (SD = 8.62) participated in the study. The intervention significantly influenced intention to extract TMs, as desired. Despite similar background and knowledge of management alternatives, participants in the intervention group had significantly lower intention to extract: control group mean (SD) = 0.39 (1.99); intervention group mean (SD) = -0.78 (1.89); mean difference (SE) = 1.17 (0.42); 95% confidence interval for the difference = 0.34 to 1.99. CONCLUSION: Results suggest this intervention, which successfully influenced a proximal predictor of behaviour pertinent to dental EBP, may result in improved EBP in a service-level trial. Basing implementation interventions and trial methodology on psychological models may effectively bridge the gap between clinical guidelines and practice.

Lessons from a randomized controlled trial designed to evaluate computer decision support software to improve the management of asthma.

PRIMARY OBJECTIVE: To investigate whether computer decision support software used in the management of patients with asthma improves clinical outcomes. RESEARCH DESIGN: Randomized controlled trial with practices each reporting on 30 patients with asthma over a 6 month period. METHODS AND PROCEDURES: 447 patients were randomly selected from practice asthma registers managed by 17 general practices from throughout the UK. Intervention practices used the software during consultations with these patients throughout the study while control practices did not. MAIN OUTCOMES AND RESULTS: Practice consultations, acute exacerbations of asthma, hospital contacts, symptoms on assessment and medication use. A smaller proportion of patients within the intervention group initiated practice consultations for their asthma: 34 (22%) vs 111 (34%), odds ratio (OR) = 0.59, 95% confidence interval (CI) (0.37-0.95); and suffered acute asthma exacerbations: 12 (8%) vs 57 (17%), OR = 0.43, 95% CI = 0.21-0.85 six months after the introduction of the computer decision support software. There were no discernable differences in reported symptoms, maintenance prescribing or use of hospital services between the two groups. CONCLUSION: The use of computer decision support software that implements guidelines during patient consultations may improve clinical outcomes for patients with asthma.

Scottish Asthma Management Initiative.

OBJECTIVES: To describe the development process of a system that links audit, research and patient care and to detail the lessons learned from establishing a Scotland wide asthma management initiative. DESIGN, SETTING AND SUBJECTS: Health Boards and practices throughout Scotland were invited to participate in an initiative which links review of care, guideline implementation, chronic disease management (CDM) approval and post-graduate education for doctors (PGEA) and nurses (PREP). Participating practices were given the materials to review 30 patients randomly selected from their asthma register. Health service resource use and drugs prescribed over a retrospective 12 month period were recorded for each patient using paper or electronic materials. All patients were invited for clinical assessment. RESULTS: A two-tier management system proved effective. Twelve of the 15 Scottish health authorities agreed to recognise the audit for automatic CDM approval although the negotiation process was prolonged; 566 practices from all parts of Scotland have expressed an interest in the initiative. Provision of distance learning material linked to PGEA accreditation is free to general practitioners (GP's) and is a useful incentive for participation. To date 42 GPs have completed the distance learning element. CONCLUSION: The Scottish Asthma Management Initiative has provided the opportunity for all sectors of the health service in Scotland to work together to explore innovative ways to improve the management and care of chronic disease. Participation in an initiative linked to guidelines, education and CDM approval is an excellent way to facilitate health professionals to improve care.

Effect of asthma and its treatment on growth: four year follow up of cohort of children from general practices in Tayside, Scotland.

OBJECTIVE: To investigate whether asthma or its treatment impairs children's growth, after allowing for socioeconomic group. DESIGN: 4 year follow up of a cohort of children aged 1-15. SETTING: 12 general practices in the Tayside region of Scotland. SUBJECTS: 3347 children with asthma or features suggestive of asthma registered with the general practices. MAIN OUTCOME MEASURES: Height and weight standard deviation scores. RESULTS: Children who lived in areas of social deprivation (assessed by postcode) had lower height and weight than their contemporaries (mean standard deviation score -0.26 (SD 1.02) and -0.18 (1.15) respectively, P < 0.001 for both). Children who were receiving > or = 400 micrograms daily of inhaled steroids and who were attending both hospital and general practice for asthma care had lower height and weight than average, independent of the effect of deprivation (mean standard deviation score -0.62 (1.01), P = 0.002, for height and -0.58 (0.94), P = 0.005, for weight). Children receiving high doses of inhaled corticosteroids also showed lower growth rates (mean change in standard deviation score -0.19 (0.51), P = 0.003). However, no other children with asthma showed growth impairment. CONCLUSION: Most children with asthma were of normal height and weight and had normal growth rates. However, children receiving high doses of inhaled steroids and requiring both general practice and hospital services had a significant reduction in their stature. This effect was independent from but smaller than the effect of socioeconomic group on stature.

Growth screening and urban deprivation.

OBJECTIVES: To assess the effect of urban deprivation on childhood growth in a modern British society by analysing data from a regional growth survey, the Tayside growth study. SETTING: The Tayside Region in Scotland, which has three districts with distinct socioeconomic status: Dundee (D, urban city), Angus (A, rural), and Perth (P, rural and county town). SUBJECTS AND METHODS: Height and weight of 23,046 children (> 90% of the regional childhood population) were measured as part of a child health surveillance programme, by community health care workers at 3, 5, 7, 9, 11, and 14 years. Height standard deviation score (calculated against Tanner) and body mass index (BMI-weight (kg)/height (m)2) were calculated for each child by a central computer program; mean height standard deviation score and BMI standard deviation score were calculated for each measuring centre (school, health clinic). A deprivation score for each centre was calculated from the prevalence of single parent families; families with more than three children; unemployment rate; the number of social class V individuals; the percentage of council houses. RESULTS: Mean height standard deviation score for Tayside was 0.11. An intraregional difference was demonstrated: mean height standard deviation score (SD) D = 0.04 (1.0); A = 0.14 (1.1); P = 0.21 (1.1); P < 0.002. There was a positive association between short stature and increasing social deprivation seen throughout Tayside (P < 0.05), with a strong association in Dundee primary school children (r = 0.6; P < 0.001). Analysis by district showed that the association was significant only above the age of 8 (P < 0.004). There was no relation between BMI and social deprivation. CONCLUSIONS: In an industrialised developed society, urban deprivation appears to influence height mostly in late childhood, and this association should be taken into consideration in the clinical management of short stature. Height seems to be a better physical indicator of urban deprivation, and hence an index of childhood health, than BMI.

Body mass index centile charts to assess fatness of British children.

Body mass index (BMI) relates weight to height and reflects the shape of a child, but because of age dependency it has not been used conventionally for the estimation of fatness in children. From measurements of Tayside children (n = 34,533) centile charts were constructed for BMI (wt/ht2) from the raw data of height and weight, using Cole's LMS method for normalised growth standards. These data were compared with the only available European BMI charts published from data of French children obtained over a period of 24 years from 1956-79. British children appear to be 'fatter'. Within a subgroup (n = 445) the BMI values were correlated with estimations of body fat, for boys and girls, from skinfold thickness (r = 0.8 and 0.81) and bioelectrical impedance (r = 0.65 and 0.7). The limits of acceptable BMI have yet to be defined.

Spectral analysis of cervical cells using the discrete Fourier transform.

This paper contains results from a preliminary study of spectral analysis techniques applied to the classification of cervical cells from routinely prepared Papanicolaou cervical smears. Experiments were conducted using a subset of 110 normals and 110 dyskaryotic single cell images randomly selected from a larger cell image data base. An assessment was made of the contribution of different regions within a cell image to the frequency spectrum. Three image sets were used, the original image itself plus two derived from it. In the first derived set, only nuclear size and shape were used. In the second set nuclear morphology and texture were included. Nuclear masking was performed using an interactive segmentation procedure. The discrete Fourier transform was applied to each image in the three image sets and classification experiments were performed using 80 features derived from the frequency spectra. An optimum set of features was selected for each experiment by canonical analysis. Good classification results were obtained when features extracted solely from nuclear shape were used. The inclusion of information relating to nuclear texture improved the results. However, inclusion of the extra nuclear region degraded the classifier's ability to discriminate between cell groups.

Automation in cervical cytology: an overview.

Cancer of the uterine cervix is one of the more common female cancers and a major source of premature female mortality. UK deaths in 1987 exceeded 2000. To reduce these rates, national screening programmes have been introduced using the Papanicolaou method. In screening for cancer, the examination of specimens is an intensive and expensive task. Its high cost has led to a number of attempts to automate the process, either fully or partially. Over the last 30 years, various experimental prescreening systems have been developed for the diagnosis of cytological samples including the use of image processing techniques. This paper presents a historical overview of automation in cervical cytology and the status of current developments for automated cytological diagnosis.

Syntax PAL: a system to improve the written syntax of language-impaired users.

In our work with children who have difficulty with spelling or with the physical action of writing, we have found a number of children who also have difficulty with written grammar. As an extension of PAL, an existing predictive spelling and typing aid, we have developed a writing aid to help these children with sentence construction. The enhanced system uses the syntax of the initial part of a sentence to enhance the position in the prediction list of syntactically correct words. It was postulated that this would discourage the use of incorrect syntax and encourage the use of correct syntax. In two case studies, the use of Syntax PAL significantly improved the quality and quantity of one child's written output, but had little effect on the other child's work.