Choosing among generalized linear models applied to medical data

Lindsey, J. K.; Jones, B.

doi:10.1002/(sici)1097-0258(19980115)17:1<59::aid-sim733>3.0.co;2-7

Cited by 152 publications

(76 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For visualisation of PTHLH expression in osteochondromas of male patients, the curve and the approximate 95% confidence intervals were estimated using P-splines [12], to model the logit of the probability. The optimal weight of the penalty was determined by Akaike Information Criterion [22]. A P value <0.05 was considered significant.…”

Section: Discussionmentioning

confidence: 99%

The use of Bcl-2 and PTHLH immunohistochemistry in the diagnosis of peripheral chondrosarcoma in a clinicopathological setting

et al. 2005

View full text Add to dashboard Cite

Distinguishing osteochondroma from low-grade secondary peripheral chondrosarcoma can be difficult. In osteochondroma, growth-signalling pathways are thought to be downregulated through exostosin (EXT) inactivation. A previous pilot study focusing on expression of putative EXT downstream effectors indicated that progression of osteochondroma towards grade I chondrosarcoma was characterised by upregulation of Bcl-2 and parathyroid hormone-like hormone (PTHLH). We investigated their use as diagnostic markers in a large nationwide series of 71 osteochondromas and 34 chondrosarcomas. Bcl-2 immunohistochemistry proved to be a valuable diagnostic tool: scoring negative in 95% (specificity) of the osteochondromas and positive in 57% (sensitivity) of the chondrosarcomas, reaching a positive predictive value of 84% and negative predictive value of 82%. Positivity was not related to age, hereditary status, gender or thickness of the cartilage cap. Presence of internal controls and verification using mRNA in situ hybridisation strengthened the reliability of the immunohistochemical staining. PTHLH showed more variable staining, being positive in osteochondromas from females or adolescent males, suggesting age- and gender-dependent expression. Thus, in cases where the distinction between osteochondroma and chondrosarcoma is difficult, Bcl-2 is a valuable diagnostic marker for malignancy, regardless of tumour size, patient gender or age, and this can be extended with PTHLH for non-adolescent male patients.

show abstract

Section: Discussionmentioning

confidence: 99%

The use of Bcl-2 and PTHLH immunohistochemistry in the diagnosis of peripheral chondrosarcoma in a clinicopathological setting

et al. 2005

View full text Add to dashboard Cite

show abstract

“…Mixed model with hospitals as random effects were used to account for the presence of possible variability between hospitals and are designed to handle nested data (like patients in each hospital) and unequal group study. 21,22 The second method analyzed differences in total costs (primary admission and Q1) among 7 categories of complications by using multivariate regression models with hospitals as random effects. First, the analysis was performed without risk-adjustment; then with risk-adjustment for patient's characteristics illustrated in Table 1; in the last analysis risk-adjustment for patient's characteristics and all type of complications were included in the mixed model.…”

Section: Statistical Analysesmentioning

confidence: 99%

Costs of complications after colorectal cancer surgery in the Netherlands: Building the business case for hospitals

Govaert

Fiocco

Dijk³

et al. 2015

European Journal of Surgical Oncology (EJSO)

View full text Add to dashboard Cite

“…As backward elimination and forward selection procedures yielded different results, the best model was found by using the Akaike information criterion [18]. Patients entering the cancer journey with any of the following of the eight people variables were statistically more likely to have at least one significant unmet need: being younger; having a long-standing illness or disability; not owning/having use of a car; not having a faith.…”

Section: Managing Emotions and Self-identitymentioning

confidence: 99%

The significant unmet needs of cancer patients: probing psychosocial concerns

Soothill

Morris

Harman

et al. 2001

Support Care Cancer

154

132

View full text Add to dashboard Cite

'Significant unmet needs' are those needs that patients identify as both important and unsatisfied. In this article we ask whether the overall needs of cancer patients are actually being met. We believe that the range of unmet need, and the kinds of patients who are more likely to claim unmet need, should be carefully identified. The needs responses of a series of 295 cancer patients in a cross-sectional survey were analysed. The majority expressed the opinion that information and good relationships with health care professionals were important, and few expressed dissatisfaction with these aspects of need. Similarly, needs items about support from family and friends were largely rated as important and satisfied. For a sizeable minority of patients, items of significant unmet need cluster around aspects of managing daily life, emotions, and social identity. The distribution of significant unmet needs is not random but is more likely to be experienced by patients who are younger, have a long-standing illness or disability, do not own/have use of a car, and/or have no religious faith. Furthermore, significant unmet needs relate to patients' ability to talk freely to a carer about the cancer, the degree to which the cancer interferes with social activities, and whether financial difficulties are experienced. Most of the significant unmet need is beyond the remit of services primarily designed for the treatment of disease. We consider whether multidisciplinary cancer teams can be expected to deal with all aspects of the cancer experience.

show abstract

Choosing among generalized linear models applied to medical data

Cited by 152 publications

References 25 publications

The use of Bcl-2 and PTHLH immunohistochemistry in the diagnosis of peripheral chondrosarcoma in a clinicopathological setting

The use of Bcl-2 and PTHLH immunohistochemistry in the diagnosis of peripheral chondrosarcoma in a clinicopathological setting

Costs of complications after colorectal cancer surgery in the Netherlands: Building the business case for hospitals

The significant unmet needs of cancer patients: probing psychosocial concerns

Contact Info

Product

Resources

About