A Scoring System for Ascertainment of Incident Stroke; the Risk Index Score (RISc)

Summary

Objectives: The main objective of this study was to develop and validate a computer-based statistical algorithm that could be translated into a simple scoring system in order to ascertain incident stroke cases using hospital admission medical records data.

Methods: The Risk Index Score (RISc) algorithm was developed using data collected prospectively by the Brain Attack Surveillance in Corpus Christi (BASIC) project, 2000. The validity of RISc was evaluated by estimating the concordance of scoring system stroke ascertainment to stroke ascertainment by physician and/or abstractor review of hospital admission records.

Results: RISc was developed on 1718 randomly selected patients (training set) and then statistically validated on an independent sample of 858 patients (validation set). A multivariable logistic model was used to develop RISc and subsequently evaluated by goodness-of-fit and receiver operating characteristic (ROC) analyses. The higher the value of RISc, the higher the patient’s risk of potential stroke. The study showed RISc was well calibrated and discriminated those who had potential stroke from those that did not on initial screening.

Conclusion: In this study we developed and validated a rapid, easy, efficient, and accurate method to ascertain incident stroke cases from routine hospital admission records for epidemiologic investigations. Validation of this scoring system was achieved statistically; however, clinical validation in a community hospital setting is warranted.

Keywords

Stroke - epidemiology - screening - surveillance - informatics

• References

• 1 Al-Wabil A, Smith MA, Moyé LA. et al. Improving efficiency of stroke research: The Brain Attack Surveillance in Corpus Christi study. J Clin Epidemiol 2003; 56: 351-7.
• 2 Klastersky J, Paesmans M, Rubenstein EB. et al. The Multinational Association for Supportive Care in Cancer Risk Index: A multinational Scoring System for Identifying Low-Risk Febrile Neutropenic Cancer Patients. J Clin Oncol 2000; 18: 3038-51.
• 3 Kass-Hout TA. Risk Index Score (RISc). Masters Abstracts International 2002. PN: 14-06487: 40: No: 2M: 440-561.
• 4 Suka M, Sugimori H, Yoshida K. Validity of the Framingham Risk Model Applied to Japanese Men. Methods Inf Med 2002; 41: 213-5.
• 5 Piriyawat P, Šmajsovà M, Smith MA. et al. Comparison of Active and Passive Surveillance for Cerebrovascular Disease: The Brain Attack Surveillance in Corpus Christi (BASIC) Project. Am J Epidemiol 2002; 156: 1062-9.
• 6 StataCorp. Stata Statistical Software. Release College Station, Texas, USA: Stata Corporation 2001
• 7 Hosmer D, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons; 2000
• 8 Cleveland WS. Robust locally weighted regression and smoothing scatterplots. J Am Stat Assoc 1979; 74: 829-36.
• 9 Sauerbrei W, Royston P. Building multivariable prognostic and diagnostic models: transformation of the predictors by using fractional polynomials. Journal of the Royal Statistical Society 1999. Series A 162
• 10 Hosmer DW, Hosmer T, le Cessie S, Lemeshow S. A comparison of goodness-of-fit tests for the logistic regression model. Stat Med 1997; 16: 965-80.
• 11 Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982; 143: 29-36.
• 12 Hajian-Tilaki KO, Hanley JA, Joseph L. et al. A comparison of parametric and nonparametric approaches to ROC analysis of quantitative diagnostic tests. Med Decis Making 1997; 17: 94-102.
• 13 Harrell FE, Lee KL, Mark DP. Multivariable Prognostic Models: Issues in Developing Models, Evaluating Assumptions and Adequacy, and Measuring and Reducing Errors. Stat Med 1996; 15 (04) 361-87.
• 14 Long SJ. Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, CA:: Sage; 1996
• 15 Dorfman DD, Alf E. Maximum likelihood estimation of parameters of single detection theory and determination of confidence intervals-rating method data. Journal of Mathematical Psychology 1969; 6: 487-96.
• 16 Picard RR, Berk KN. Data Splitting. American Statistician 1990; 44: 140-7.
• 17 Efron B, Tibshirani R. Cross-validation and the bootstrap: Estimating the error rate of a prediction rule. Technical Report TR-477, Dept. of Statistics, Stanford University 1995
• 18 DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the area under two or more correlated receiver operating curves: A nonparametric approach. Biometrics 1988; 44: 837-45.
• 19 Steyerberg EW, Borsboom GJJM, van Houwelingen HC, Eijkemans MJC, Habbema JDF. Validation and updating of predictive logistic regression models: a study on sample size and shrinkage. Stat Med 2004; 23: 2567-86.
• 20 Schwarzer G, Nagata T, Mattern D, Schmelzeisen R, Schumacher M. Comparison of Fuzzy Inference, Logistic Regression, and Classification Trees (CART) Prediction of Cervical Lymph Node Metastasis in Carcinoma of the Tongue. Methods Inf Med 2003; 42: 572-7.