Performance of a statistical model to predict stroke outcome in the context of a large, simple, randomized, controlled trial of feeding

M. S. Dennis, G. Cranswick, A. Fraser, S. Grant, A. Gunkel, J. Hunter, S. Lewis, D. Perry, V. Soosay, A. Williamson, A. Young, J. Bulpitt, A. Grant, G. Murray, P. Sandercock, N. Anderson, S. Bahar, G. Hankey, S. Ricci, G. Bathgate & 79 others C. Chalmers, B. Farrell, J. Forbes, S. Ghosh, P. Langhorne, J. MacIntyre, C. A. McAteer, P. O'Neill, J. Potter, M. Roberts, C. Warlow, C. Counsel, G. J. Hankey, Stephen McDonald, T. Bennett, J. Karrasch, C. Lowe, E. O'Brien, F. Simpson, A. Bramley, J. Frayne, G. Baker, G. DeGabriele, J. Kennett, J. Nevin, P. D. Aitken, G. Hall, V. Schotte, C. Vandenbruaene, G. T.O. Vanhooren, C. Vanmaele, C. Andre, M. A.S.D. Lima, M. O. Py, S. Alward, P. Bailey, P. Cook, S. J. Phillips, Y. Reidy, E. Ehler, P. Geier, P. Vyhnalek, C. Majvald, D. Skoloudik, L. Bech, D. Rizzi, T. Soerensen, D. Rasmussen, K. Y. Chan, E. S.L. Chow, C. K.L. Kng, C. P. Wong, U. Devraj, Manjari, J. L. Pinheiro, A. K. Roy, S. Panda, K. Prasad, M. Tripathi, L. Ambrosius, G. Benemio, M. G. Celani, C. Ottaviani, B. Randolph, E. Righetti, A. Tufi, D. Sita, P. Vanni, G. Volpi, M. Del Gobbo, O. Scarpino, G. Benati, V. Pedone, A. Ciccone, I. Santilli, R. Sterzi, O. Cazzato, P. Parise, A. G. Gregoris

Research output: Contribution to journalArticle

35 Citations (Scopus)

Abstract

Background and Purpose - Statistical models to predict the outcome of stroke patients have several uses. Their utility depends on their predictive accuracy in patients other than those on whom they were developed (ie, external validity). We sought to test the external validity of some recently described models in patients enrolled in the FOOD (Feed Or Ordinary Diet) trial: a large randomized trial evaluating feeding policies in patients with stroke. Methods - The predictive variables were collected during a telephone call to randomize the patient a median of 5 days after stroke onset. Patients were followed up 6 months later to establish their survival, functional status, and residence. Charts were plotted to demonstrate the discrimination and calibration of the models. Results - The models performed well in the first 2955 patients enrolled and followed up in the FOOD trial. The area under the receiver operating characteristic curves varied between 0.78 and 0.81 (with 0.5 indicating no discrimination and 1.0 indicating perfect discrimination). The discrimination was marginally better for patients enrolled within the first day of stroke than later. The models tended to provide rather pessimistic predictions in all groups except those predicted to have a high likelihood of surviving free of dependency. Conclusions - As one might predict, the discriminatory power in the selected cohort of trial patients was marginally less good than in previously studied unselected cohorts used to test their external validity. These models provide a well-tested tool for stratification in trials, comparing outcomes in different cohorts and examining the additional predictive power of novel factors.

LanguageEnglish
Pages127-133
Number of pages7
JournalStroke
Volume34
Issue number1
DOIs
Publication statusPublished - 1 Jan 2003

Keywords

  • Cerebrovascular disorders
  • Outcome
  • Prognosis

ASJC Scopus subject areas

  • Clinical Neurology
  • Cardiology and Cardiovascular Medicine
  • Advanced and Specialised Nursing

Cite this

Dennis, M. S. ; Cranswick, G. ; Fraser, A. ; Grant, S. ; Gunkel, A. ; Hunter, J. ; Lewis, S. ; Perry, D. ; Soosay, V. ; Williamson, A. ; Young, A. ; Bulpitt, J. ; Grant, A. ; Murray, G. ; Sandercock, P. ; Anderson, N. ; Bahar, S. ; Hankey, G. ; Ricci, S. ; Bathgate, G. ; Chalmers, C. ; Farrell, B. ; Forbes, J. ; Ghosh, S. ; Langhorne, P. ; MacIntyre, J. ; McAteer, C. A. ; O'Neill, P. ; Potter, J. ; Roberts, M. ; Warlow, C. ; Counsel, C. ; Hankey, G. J. ; McDonald, Stephen ; Bennett, T. ; Karrasch, J. ; Lowe, C. ; O'Brien, E. ; Simpson, F. ; Bramley, A. ; Frayne, J. ; Baker, G. ; DeGabriele, G. ; Kennett, J. ; Nevin, J. ; Aitken, P. D. ; Hall, G. ; Schotte, V. ; Vandenbruaene, C. ; Vanhooren, G. T.O. ; Vanmaele, C. ; Andre, C. ; Lima, M. A.S.D. ; Py, M. O. ; Alward, S. ; Bailey, P. ; Cook, P. ; Phillips, S. J. ; Reidy, Y. ; Ehler, E. ; Geier, P. ; Vyhnalek, P. ; Majvald, C. ; Skoloudik, D. ; Bech, L. ; Rizzi, D. ; Soerensen, T. ; Rasmussen, D. ; Chan, K. Y. ; Chow, E. S.L. ; Kng, C. K.L. ; Wong, C. P. ; Devraj, U. ; Manjari ; Pinheiro, J. L. ; Roy, A. K. ; Panda, S. ; Prasad, K. ; Tripathi, M. ; Ambrosius, L. ; Benemio, G. ; Celani, M. G. ; Ottaviani, C. ; Randolph, B. ; Righetti, E. ; Tufi, A. ; Sita, D. ; Vanni, P. ; Volpi, G. ; Del Gobbo, M. ; Scarpino, O. ; Benati, G. ; Pedone, V. ; Ciccone, A. ; Santilli, I. ; Sterzi, R. ; Cazzato, O. ; Parise, P. ; Gregoris, A. G. / Performance of a statistical model to predict stroke outcome in the context of a large, simple, randomized, controlled trial of feeding. In: Stroke. 2003 ; Vol. 34, No. 1. pp. 127-133.
@article{9a612d76dbac437f92ee0e9164ec97c5,
title = "Performance of a statistical model to predict stroke outcome in the context of a large, simple, randomized, controlled trial of feeding",
abstract = "Background and Purpose - Statistical models to predict the outcome of stroke patients have several uses. Their utility depends on their predictive accuracy in patients other than those on whom they were developed (ie, external validity). We sought to test the external validity of some recently described models in patients enrolled in the FOOD (Feed Or Ordinary Diet) trial: a large randomized trial evaluating feeding policies in patients with stroke. Methods - The predictive variables were collected during a telephone call to randomize the patient a median of 5 days after stroke onset. Patients were followed up 6 months later to establish their survival, functional status, and residence. Charts were plotted to demonstrate the discrimination and calibration of the models. Results - The models performed well in the first 2955 patients enrolled and followed up in the FOOD trial. The area under the receiver operating characteristic curves varied between 0.78 and 0.81 (with 0.5 indicating no discrimination and 1.0 indicating perfect discrimination). The discrimination was marginally better for patients enrolled within the first day of stroke than later. The models tended to provide rather pessimistic predictions in all groups except those predicted to have a high likelihood of surviving free of dependency. Conclusions - As one might predict, the discriminatory power in the selected cohort of trial patients was marginally less good than in previously studied unselected cohorts used to test their external validity. These models provide a well-tested tool for stratification in trials, comparing outcomes in different cohorts and examining the additional predictive power of novel factors.",
keywords = "Cerebrovascular disorders, Outcome, Prognosis",
author = "Dennis, {M. S.} and G. Cranswick and A. Fraser and S. Grant and A. Gunkel and J. Hunter and S. Lewis and D. Perry and V. Soosay and A. Williamson and A. Young and J. Bulpitt and A. Grant and G. Murray and P. Sandercock and N. Anderson and S. Bahar and G. Hankey and S. Ricci and G. Bathgate and C. Chalmers and B. Farrell and J. Forbes and S. Ghosh and P. Langhorne and J. MacIntyre and McAteer, {C. A.} and P. O'Neill and J. Potter and M. Roberts and C. Warlow and C. Counsel and Hankey, {G. J.} and Stephen McDonald and T. Bennett and J. Karrasch and C. Lowe and E. O'Brien and F. Simpson and A. Bramley and J. Frayne and G. Baker and G. DeGabriele and J. Kennett and J. Nevin and Aitken, {P. D.} and G. Hall and V. Schotte and C. Vandenbruaene and Vanhooren, {G. T.O.} and C. Vanmaele and C. Andre and Lima, {M. A.S.D.} and Py, {M. O.} and S. Alward and P. Bailey and P. Cook and Phillips, {S. J.} and Y. Reidy and E. Ehler and P. Geier and P. Vyhnalek and C. Majvald and D. Skoloudik and L. Bech and D. Rizzi and T. Soerensen and D. Rasmussen and Chan, {K. Y.} and Chow, {E. S.L.} and Kng, {C. K.L.} and Wong, {C. P.} and U. Devraj and Manjari and Pinheiro, {J. L.} and Roy, {A. K.} and S. Panda and K. Prasad and M. Tripathi and L. Ambrosius and G. Benemio and Celani, {M. G.} and C. Ottaviani and B. Randolph and E. Righetti and A. Tufi and D. Sita and P. Vanni and G. Volpi and {Del Gobbo}, M. and O. Scarpino and G. Benati and V. Pedone and A. Ciccone and I. Santilli and R. Sterzi and O. Cazzato and P. Parise and Gregoris, {A. G.}",
year = "2003",
month = "1",
day = "1",
doi = "10.1161/01.STR.0000044165.41303.50",
language = "English",
volume = "34",
pages = "127--133",
journal = "Stroke",
issn = "0039-2499",
publisher = "Lippincott Williams and Wilkins",
number = "1",

}

Dennis, MS, Cranswick, G, Fraser, A, Grant, S, Gunkel, A, Hunter, J, Lewis, S, Perry, D, Soosay, V, Williamson, A, Young, A, Bulpitt, J, Grant, A, Murray, G, Sandercock, P, Anderson, N, Bahar, S, Hankey, G, Ricci, S, Bathgate, G, Chalmers, C, Farrell, B, Forbes, J, Ghosh, S, Langhorne, P, MacIntyre, J, McAteer, CA, O'Neill, P, Potter, J, Roberts, M, Warlow, C, Counsel, C, Hankey, GJ, McDonald, S, Bennett, T, Karrasch, J, Lowe, C, O'Brien, E, Simpson, F, Bramley, A, Frayne, J, Baker, G, DeGabriele, G, Kennett, J, Nevin, J, Aitken, PD, Hall, G, Schotte, V, Vandenbruaene, C, Vanhooren, GTO, Vanmaele, C, Andre, C, Lima, MASD, Py, MO, Alward, S, Bailey, P, Cook, P, Phillips, SJ, Reidy, Y, Ehler, E, Geier, P, Vyhnalek, P, Majvald, C, Skoloudik, D, Bech, L, Rizzi, D, Soerensen, T, Rasmussen, D, Chan, KY, Chow, ESL, Kng, CKL, Wong, CP, Devraj, U, Manjari, Pinheiro, JL, Roy, AK, Panda, S, Prasad, K, Tripathi, M, Ambrosius, L, Benemio, G, Celani, MG, Ottaviani, C, Randolph, B, Righetti, E, Tufi, A, Sita, D, Vanni, P, Volpi, G, Del Gobbo, M, Scarpino, O, Benati, G, Pedone, V, Ciccone, A, Santilli, I, Sterzi, R, Cazzato, O, Parise, P & Gregoris, AG 2003, 'Performance of a statistical model to predict stroke outcome in the context of a large, simple, randomized, controlled trial of feeding', Stroke, vol. 34, no. 1, pp. 127-133. https://doi.org/10.1161/01.STR.0000044165.41303.50

Performance of a statistical model to predict stroke outcome in the context of a large, simple, randomized, controlled trial of feeding. / Dennis, M. S.; Cranswick, G.; Fraser, A.; Grant, S.; Gunkel, A.; Hunter, J.; Lewis, S.; Perry, D.; Soosay, V.; Williamson, A.; Young, A.; Bulpitt, J.; Grant, A.; Murray, G.; Sandercock, P.; Anderson, N.; Bahar, S.; Hankey, G.; Ricci, S.; Bathgate, G.; Chalmers, C.; Farrell, B.; Forbes, J.; Ghosh, S.; Langhorne, P.; MacIntyre, J.; McAteer, C. A.; O'Neill, P.; Potter, J.; Roberts, M.; Warlow, C.; Counsel, C.; Hankey, G. J.; McDonald, Stephen; Bennett, T.; Karrasch, J.; Lowe, C.; O'Brien, E.; Simpson, F.; Bramley, A.; Frayne, J.; Baker, G.; DeGabriele, G.; Kennett, J.; Nevin, J.; Aitken, P. D.; Hall, G.; Schotte, V.; Vandenbruaene, C.; Vanhooren, G. T.O.; Vanmaele, C.; Andre, C.; Lima, M. A.S.D.; Py, M. O.; Alward, S.; Bailey, P.; Cook, P.; Phillips, S. J.; Reidy, Y.; Ehler, E.; Geier, P.; Vyhnalek, P.; Majvald, C.; Skoloudik, D.; Bech, L.; Rizzi, D.; Soerensen, T.; Rasmussen, D.; Chan, K. Y.; Chow, E. S.L.; Kng, C. K.L.; Wong, C. P.; Devraj, U.; Manjari; Pinheiro, J. L.; Roy, A. K.; Panda, S.; Prasad, K.; Tripathi, M.; Ambrosius, L.; Benemio, G.; Celani, M. G.; Ottaviani, C.; Randolph, B.; Righetti, E.; Tufi, A.; Sita, D.; Vanni, P.; Volpi, G.; Del Gobbo, M.; Scarpino, O.; Benati, G.; Pedone, V.; Ciccone, A.; Santilli, I.; Sterzi, R.; Cazzato, O.; Parise, P.; Gregoris, A. G.

In: Stroke, Vol. 34, No. 1, 01.01.2003, p. 127-133.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Performance of a statistical model to predict stroke outcome in the context of a large, simple, randomized, controlled trial of feeding

AU - Dennis, M. S.

AU - Cranswick, G.

AU - Fraser, A.

AU - Grant, S.

AU - Gunkel, A.

AU - Hunter, J.

AU - Lewis, S.

AU - Perry, D.

AU - Soosay, V.

AU - Williamson, A.

AU - Young, A.

AU - Bulpitt, J.

AU - Grant, A.

AU - Murray, G.

AU - Sandercock, P.

AU - Anderson, N.

AU - Bahar, S.

AU - Hankey, G.

AU - Ricci, S.

AU - Bathgate, G.

AU - Chalmers, C.

AU - Farrell, B.

AU - Forbes, J.

AU - Ghosh, S.

AU - Langhorne, P.

AU - MacIntyre, J.

AU - McAteer, C. A.

AU - O'Neill, P.

AU - Potter, J.

AU - Roberts, M.

AU - Warlow, C.

AU - Counsel, C.

AU - Hankey, G. J.

AU - McDonald, Stephen

AU - Bennett, T.

AU - Karrasch, J.

AU - Lowe, C.

AU - O'Brien, E.

AU - Simpson, F.

AU - Bramley, A.

AU - Frayne, J.

AU - Baker, G.

AU - DeGabriele, G.

AU - Kennett, J.

AU - Nevin, J.

AU - Aitken, P. D.

AU - Hall, G.

AU - Schotte, V.

AU - Vandenbruaene, C.

AU - Vanhooren, G. T.O.

AU - Vanmaele, C.

AU - Andre, C.

AU - Lima, M. A.S.D.

AU - Py, M. O.

AU - Alward, S.

AU - Bailey, P.

AU - Cook, P.

AU - Phillips, S. J.

AU - Reidy, Y.

AU - Ehler, E.

AU - Geier, P.

AU - Vyhnalek, P.

AU - Majvald, C.

AU - Skoloudik, D.

AU - Bech, L.

AU - Rizzi, D.

AU - Soerensen, T.

AU - Rasmussen, D.

AU - Chan, K. Y.

AU - Chow, E. S.L.

AU - Kng, C. K.L.

AU - Wong, C. P.

AU - Devraj, U.

AU - Manjari,

AU - Pinheiro, J. L.

AU - Roy, A. K.

AU - Panda, S.

AU - Prasad, K.

AU - Tripathi, M.

AU - Ambrosius, L.

AU - Benemio, G.

AU - Celani, M. G.

AU - Ottaviani, C.

AU - Randolph, B.

AU - Righetti, E.

AU - Tufi, A.

AU - Sita, D.

AU - Vanni, P.

AU - Volpi, G.

AU - Del Gobbo, M.

AU - Scarpino, O.

AU - Benati, G.

AU - Pedone, V.

AU - Ciccone, A.

AU - Santilli, I.

AU - Sterzi, R.

AU - Cazzato, O.

AU - Parise, P.

AU - Gregoris, A. G.

PY - 2003/1/1

Y1 - 2003/1/1

N2 - Background and Purpose - Statistical models to predict the outcome of stroke patients have several uses. Their utility depends on their predictive accuracy in patients other than those on whom they were developed (ie, external validity). We sought to test the external validity of some recently described models in patients enrolled in the FOOD (Feed Or Ordinary Diet) trial: a large randomized trial evaluating feeding policies in patients with stroke. Methods - The predictive variables were collected during a telephone call to randomize the patient a median of 5 days after stroke onset. Patients were followed up 6 months later to establish their survival, functional status, and residence. Charts were plotted to demonstrate the discrimination and calibration of the models. Results - The models performed well in the first 2955 patients enrolled and followed up in the FOOD trial. The area under the receiver operating characteristic curves varied between 0.78 and 0.81 (with 0.5 indicating no discrimination and 1.0 indicating perfect discrimination). The discrimination was marginally better for patients enrolled within the first day of stroke than later. The models tended to provide rather pessimistic predictions in all groups except those predicted to have a high likelihood of surviving free of dependency. Conclusions - As one might predict, the discriminatory power in the selected cohort of trial patients was marginally less good than in previously studied unselected cohorts used to test their external validity. These models provide a well-tested tool for stratification in trials, comparing outcomes in different cohorts and examining the additional predictive power of novel factors.

AB - Background and Purpose - Statistical models to predict the outcome of stroke patients have several uses. Their utility depends on their predictive accuracy in patients other than those on whom they were developed (ie, external validity). We sought to test the external validity of some recently described models in patients enrolled in the FOOD (Feed Or Ordinary Diet) trial: a large randomized trial evaluating feeding policies in patients with stroke. Methods - The predictive variables were collected during a telephone call to randomize the patient a median of 5 days after stroke onset. Patients were followed up 6 months later to establish their survival, functional status, and residence. Charts were plotted to demonstrate the discrimination and calibration of the models. Results - The models performed well in the first 2955 patients enrolled and followed up in the FOOD trial. The area under the receiver operating characteristic curves varied between 0.78 and 0.81 (with 0.5 indicating no discrimination and 1.0 indicating perfect discrimination). The discrimination was marginally better for patients enrolled within the first day of stroke than later. The models tended to provide rather pessimistic predictions in all groups except those predicted to have a high likelihood of surviving free of dependency. Conclusions - As one might predict, the discriminatory power in the selected cohort of trial patients was marginally less good than in previously studied unselected cohorts used to test their external validity. These models provide a well-tested tool for stratification in trials, comparing outcomes in different cohorts and examining the additional predictive power of novel factors.

KW - Cerebrovascular disorders

KW - Outcome

KW - Prognosis

UR - http://www.scopus.com/inward/record.url?scp=0037225107&partnerID=8YFLogxK

U2 - 10.1161/01.STR.0000044165.41303.50

DO - 10.1161/01.STR.0000044165.41303.50

M3 - Article

VL - 34

SP - 127

EP - 133

JO - Stroke

T2 - Stroke

JF - Stroke

SN - 0039-2499

IS - 1

ER -