Ces Urol 1999, 3(1):43-47
Mathematic simulation of illness prognosis in curable renal carcinoma
- Katedra pravděpodobnosti a matematické statistiky MFF UK Praha Urologická klinika 2. IF UK Praha a FN Motol
This study was conducted to develop a prognostic model for predicting 5-year crude survival for patients with renal cell carcinoma. The sample population was comprised of 121 patients treated with radical nephrectomy in Faculty hospital Prague - Motol between 1982 - 1993. Patients with inoperable tumour or distant metastases were excluded. The overall 5-year survival rate of the group was 68,6 % (83/121). The significance of age, sex, stage, grade and lymph node involvement as prognostic factors was evaluated. Data analysis and prognostic model construction were performed using logistic regression techniques. The statistical analysis was conducted using software packages Matlab and S-plus.
Prognostic factors were ordered according to statistical significance as follows: age, grade, stage, sex and lymph node involvement. The highest impact on 5-year survival had grade G3 (the odds ratio 5,0 between G2 and G3) and tumour extension through the renal capsule (difference between pT2 and pT3a), where the odds ratio was 3,6.
To asses the correctness of the prognostic model by using the Hosmer and Lemeshow test, the data was split into a model-building group (1982-1991) and a validation group (1992-1993). Using only the model-building group a prognostic model was fitted to the data. Evaluation of the goodness of fit of the model yielded a P value 0.389, respectively 0.153 in the model-building, respectively in the validation group. Both P values are higher then 0.05, which indicates no evidence of lack of fit.
The prognosis of patients with renal cell carcinoma, calculated using the estimated mathematical model, does not differ statistically significantly from the real 5-year survival.
Keywords: Renal cell carcinoma, Prognostic factors, Mathematic model of prognosis
Published: January 1, 1999
References
- Harrel, F.E., Lee, K.L., Mark, D.B.: Tutorial in Biostatistics. Multivariable Prognostic Models: Issues in Developing Models, Evaluating Assumptions and Adequacy, and Measuring and Reducing Errors, Statistics in Medicine, 1996, 15, s. 361-387
Go to original source... - Wyatt, J.C., Altman, D.G.: Prognostic models: clinically useful or quickly forgotten? British Medical Journal, 1995, 311, s. 1539-1541 Go to original source...
- Hosmer, D.W., Lemeshow, S.: Model-Building Strategies and Methods for Logistic Regression, In: Applied Logistic Regression, New York: Wiley, 1989 Go to original source...
- Hosmer, D.W., Lemeshow, S.: Assessing the Fit of the Model, In: Applied Logistic Regression, New York: Wiley, 1989
- Kawaciuk, I., Dušek, P., Hanek, P.: Pětileté přežití nemocných s karcinomem ledviny v závislosti na stagingu a gradingu nádoru, Česká Urologie, 1, 1998, s. 4-7