Grading gems: Appraising the quality of research about social work and social care.

The impetus towards basing practice and policy decisions more explicitly on sound research requires tools to facilitate the systematic appraisal of the quality of research encompassing a diverse range of methods and designs. Five exemplar tools were developed and assessed in terms of their usefulness in selecting studies for inclusion in a systematic review. The widely used ‘hierarchy of evidence’ was adapted and used to appraise internal validity. Four tools were then developed to appraise the external validity dimensions of generalizability (two scales) and methods of data collection (two scales). Methods of combining the scores generated by each tool were explored. Qualitative and quantitative studies were appraised, not separated into two spheres but by using complementary tools developed to appraise different aspects of rigour. There was a high level of agreement between researchers in applying the tools to twenty-two studies on decision making by professionals about the longterm care of older people. The scales for internal validity and generalizability discriminated between the qualities of studies appropriately. The two tools to appraise data collection gave diverse results. Excluding studies that scored in the lowest category on any scale appeared to be the scoring system that was most justifiable. This approach is presented to stimulate debate about the practical application of the evidence-based initiative to social work and social care. This study may assist in developing clearer definitions and common language about appraising rigour that should further the process of selecting robust research for synthesis to inform practice and policy decisions.

Interobserver agreement of the Gross Motor Function Classification System in an ambulant population of children with cerebral palsy

Gross Motor Function Classification System (GMFCS) level was reported by three independent assessors in a population of children with cerebral palsy (CP) aged between 4 and 18 years (n=184; 112 males, 72 females; mean age 10y 10mo [SD 3y 7mo]). A software algorithm also provided a computed GMFCS level from a regional CP registry. Participants had clinical diagnoses of unilateral (n=94) and bilateral (n=84) spastic CP, ataxia (n=4), dyskinesia (n=1), and hypotonia (n=1), and could walk independently with or without the use of an aid (GMFCS Levels I-IV). Research physiotherapist (n=184) and parent/guardian data (n=178) were collected in a research environment. Data from the child's community physiotherapist (n=143) were obtained by postal questionnaire. Results, using the kappa statistic with linear weighting (?1w), showed good agreement between the parent/guardian and research physiotherapist (?1w=0.75) with more moderate levels of agreement between the clinical physiotherapist and researcher (?1w=0.64) and the clinical physiotherapist and parent/guardian (?1w=0.57). Agreement was consistently better for older children (>2y). This study has shown that agreement with parent report increases with therapists'experience of the GMFCS and knowledge of the child at the time of grading. Substantial agreement between a computed GMFCS and an experienced therapist (?1w=0.74) also demonstrates the potential for extrapolation of GMFCS rating from an existing CP registry, providing the latter has sufficient data on locomotor ability.

PROSTATIC INTRAEPITHELIAL NEOPLASIA (PIN) - PERFORMANCE OF BAYESIAN BELIEF NETWORK FOR DIAGNOSIS AND GRADING

Prostatic intraepithelial neoplasia (PIN) diagnosis and grading are affected by uncertainties which arise from the fact that almost all knowledge of PIN histopathology is expressed in concepts, descriptive linguistic terms, and words. A Bayesian belief network (BBN) was therefore used to reduce the problem of uncertainty in diagnostic clue assessment, while still considering the dependences between elements in the reasoning sequence. A shallow network was used with an open-tree topology, with eight first-level descendant nodes for the diagnostic clues (evidence nodes), each independently linked by a conditional probability matrix to a root node containing the diagnostic alternatives (decision node). One of the evidence nodes was based on the tissue architecture and the others were based on cell features. The system was designed to be interactive, in that the histopathologist entered evidence into the network in the form of likelihood ratios for outcomes at each evidence node. The efficiency of the network was tested on a series of 110 prostate specimens, subdivided as follows: 22 cases of non-neoplastic prostate or benign prostatic tissue (NP), 22 PINs of low grade (PINlow), 22 PINs of high grade (PINhigh), 22 prostatic adenocarcinomas with cribriform pattern (PACcri), and 22 prostatic adenocarcinomas with large acinar pattern (PAClgac). The results obtained in the benign and malignant categories showed that the belief for the diagnostic alternatives is very high, the values being in general more than 0.8 and often close to 1.0. When considering the PIN lesions, the network classified and graded most of the cases with high certainty. However, there were some cases which showed values less than 0.8 (13 cases out of 44), thus indicating that there are situations in which the feature changes are intermediate between contiguous categories or grades. Discrepancy between morphological grading and the BBN results was observed in four out of 44 PIN cases: one PINlow was classified as PINhigh and three PINhigh were classified as PINlow. In conclusion, the network can grade PlN lesions and differentiate them from other prostate lesions with certainty. In particular, it offers a descriptive classifier which is readily implemented and which allows the use of linguistic, fuzzy variables.