Predicting 30-Day Hospital Readmission Risk in a National Cohort of Patients with Cirrhosis.
Calibration
Cirrhosis
Hospital readmission
Logistic regression
Risk prediction
Journal
Digestive diseases and sciences
ISSN: 1573-2568
Titre abrégé: Dig Dis Sci
Pays: United States
ID NLM: 7902782
Informations de publication
Date de publication:
04 2020
04 2020
Historique:
received:
25
04
2019
accepted:
04
09
2019
pubmed:
19
9
2019
medline:
6
8
2020
entrez:
19
9
2019
Statut:
ppublish
Résumé
Early hospital readmission for patients with cirrhosis continues to challenge the healthcare system. Risk stratification may help tailor resources, but existing models were designed using small, single-institution cohorts or had modest performance. We leveraged a large clinical database from the Department of Veterans Affairs (VA) to design a readmission risk model for patients hospitalized with cirrhosis. Additionally, we analyzed potentially modifiable or unexplored readmission risk factors. A national VA retrospective cohort of patients with a history of cirrhosis hospitalized for any reason from January 1, 2006, to November 30, 2013, was developed from 123 centers. Using 174 candidate variables within demographics, laboratory results, vital signs, medications, diagnoses and procedures, and healthcare utilization, we built a 47-variable penalized logistic regression model with the outcome of all-cause 30-day readmission. We excluded patients who left against medical advice, transferred to a non-VA facility, or if the hospital length of stay was greater than 30 days. We evaluated calibration and discrimination across variable volume and compared the performance to recalibrated preexisting risk models for readmission. We analyzed 67,749 patients and 179,298 index hospitalizations. The 30-day readmission rate was 23%. Ascites was the most common cirrhosis-related cause of index hospitalization and readmission. The AUC of the model was 0.670 compared to existing models (0.649, 0.566, 0.577). The Brier score of 0.165 showed good calibration. Our model achieved better discrimination and calibration compared to existing models, even after local recalibration. Assessment of calibration by variable parsimony revealed performance improvements for increasing variable inclusion well beyond those detectable for discrimination.
Sections du résumé
BACKGROUND
Early hospital readmission for patients with cirrhosis continues to challenge the healthcare system. Risk stratification may help tailor resources, but existing models were designed using small, single-institution cohorts or had modest performance.
AIMS
We leveraged a large clinical database from the Department of Veterans Affairs (VA) to design a readmission risk model for patients hospitalized with cirrhosis. Additionally, we analyzed potentially modifiable or unexplored readmission risk factors.
METHODS
A national VA retrospective cohort of patients with a history of cirrhosis hospitalized for any reason from January 1, 2006, to November 30, 2013, was developed from 123 centers. Using 174 candidate variables within demographics, laboratory results, vital signs, medications, diagnoses and procedures, and healthcare utilization, we built a 47-variable penalized logistic regression model with the outcome of all-cause 30-day readmission. We excluded patients who left against medical advice, transferred to a non-VA facility, or if the hospital length of stay was greater than 30 days. We evaluated calibration and discrimination across variable volume and compared the performance to recalibrated preexisting risk models for readmission.
RESULTS
We analyzed 67,749 patients and 179,298 index hospitalizations. The 30-day readmission rate was 23%. Ascites was the most common cirrhosis-related cause of index hospitalization and readmission. The AUC of the model was 0.670 compared to existing models (0.649, 0.566, 0.577). The Brier score of 0.165 showed good calibration.
CONCLUSION
Our model achieved better discrimination and calibration compared to existing models, even after local recalibration. Assessment of calibration by variable parsimony revealed performance improvements for increasing variable inclusion well beyond those detectable for discrimination.
Identifiants
pubmed: 31531817
doi: 10.1007/s10620-019-05826-w
pii: 10.1007/s10620-019-05826-w
pmc: PMC7073276
mid: NIHMS1540185
doi:
Types de publication
Journal Article
Research Support, N.I.H., Extramural
Research Support, U.S. Gov't, Non-P.H.S.
Langues
eng
Sous-ensembles de citation
IM
Pagination
1003-1031Subventions
Organisme : NIH HHS
ID : U2C OD023196
Pays : United States
Organisme : NLM NIH HHS
ID : 5T15LM007450
Pays : United States
Organisme : HSRD VA
ID : I01 HX001284
Pays : United States
Organisme : NIH HHS
ID : 1U2COD023196
Pays : United States
Organisme : EPA
ID : EP-D-13-052
Pays : United States
Références
Int J Med Inform. 2003 Mar;69(2-3):135-56
pubmed: 12810119
Gastroenterology. 2012 Jul;143(1):70-7
pubmed: 22465432
Stat Med. 2000 Apr 30;19(8):1059-79
pubmed: 10790680
Health Aff (Millwood). 2014 Jul;33(7):1203-11
pubmed: 25006147
Lancet. 2008 Mar 8;371(9615):838-51
pubmed: 18328931
Health Aff (Millwood). 2014 Jul;33(7):1148-54
pubmed: 25006140
Adv Ther. 2013 Jan;30(1):71-80
pubmed: 23292659
J Hosp Med. 2015 Apr;10(4):236-41
pubmed: 25557938
Am J Gastroenterol. 2016 Jan;111(1):87-92
pubmed: 26729545
Clin Gastroenterol Hepatol. 2011 Jun;9(6):524-530.e1; quiz e60
pubmed: 21440669
Annu Rev Biomed Eng. 2006;8:567-99
pubmed: 16834567
Clin Liver Dis (Hoboken). 2015 Jan 20;4(6):138-140
pubmed: 30992941
BMJ Qual Saf. 2013 Dec;22(12):998-1005
pubmed: 23904506
Drug Alcohol Depend. 2011 Jul 1;116(1-3):93-101
pubmed: 21277712
Med Care. 2006 Nov;44(11):972-81
pubmed: 17063128
Nature. 1999 Oct 21;401(6755):788-91
pubmed: 10548103
J Hosp Med. 2009 Apr;4(4):211-8
pubmed: 19388074
Stat Med. 2008 Jan 30;27(2):157-72; discussion 207-12
pubmed: 17569110
J Biomed Inform. 2015 Apr;54:283-93
pubmed: 25579635
Clin Gastroenterol Hepatol. 2013 Oct;11(10):1335-1341.e1
pubmed: 23591286
JAMA. 2016 Feb 16;315(7):651-2
pubmed: 26881365
Clin Gastroenterol Hepatol. 2011 Mar;9(3):254-9
pubmed: 21092762
Clin Orthop Relat Res. 2010 Jan;468(1):57-63
pubmed: 19844772
Hepatology. 2002 Jul;36(1):227-42
pubmed: 12085369
Gastroenterology. 2015 Nov;149(6):1471-1482.e5; quiz e17-8
pubmed: 26255044
Eur J Cardiothorac Surg. 2013 Jun;43(6):1146-52
pubmed: 23152436
Dig Dis Sci. 2003 Aug;48(8):1622-6
pubmed: 12924658
BMJ. 2009 Jun 04;338:b606
pubmed: 19502216
J Clin Epidemiol. 2016 Jun;74:167-76
pubmed: 26772608
Stat Med. 1996 Feb 28;15(4):361-87
pubmed: 8668867
J Hosp Med. 2018 Apr 25;:
pubmed: 29694458
Med Decis Making. 2012 May-Jun;32(3):E1-10
pubmed: 22427369
Dig Liver Dis. 2017 Aug;49(8):903-909
pubmed: 28410915
Clin Sci (Lond). 2011 Dec;121(11):509-21
pubmed: 21692745
Heart. 2012 May;98(9):691-8
pubmed: 22397946
Med Decis Making. 2015 Feb;35(2):162-9
pubmed: 25155798
Epidemiology. 2014 Jan;25(1):114-21
pubmed: 24240655
Gastroenterology. 2019 Jan;156(1):254-272.e11
pubmed: 30315778
AMIA Annu Symp Proc. 2014 Nov 14;2014:424-31
pubmed: 25954346
JAMA Pediatr. 2017 Apr 1;171(4):365-371
pubmed: 28241253
JAMA. 2016 Apr 26;315(16):1713-4
pubmed: 27115375
J Clin Epidemiol. 2008 Nov;61(11):1085-94
pubmed: 19208371
Clin Gastroenterol Hepatol. 2016 May;14(5):753-9
pubmed: 26407750
Hepatology. 2012 Jan;55(1):184-91
pubmed: 21858847
Gastroenterology. 2012 Nov;143(5):1179-1187.e3
pubmed: 22885331
J Pain Symptom Manage. 2016 Sep;52(3):412-419.e1
pubmed: 27265812
Gastroenterology. 2009 Apr;136(4):1134-44
pubmed: 19245868
Med Care. 2005 Nov;43(11):1130-9
pubmed: 16224307
Am J Gastroenterol. 2011 Sep;106(9):1646-53
pubmed: 21556040
J Hepatol. 2013 Aug;59(2):257-64
pubmed: 23523582
Stat Med. 2014 Jun 30;33(14):2390-407
pubmed: 24497413
Gastroenterology. 2013 Aug;145(2):375-82.e1-2
pubmed: 23583430
J Clin Gastroenterol. 2013 May-Jun;47(5):e50-4
pubmed: 23090041
CMAJ. 2011 Apr 19;183(7):E391-402
pubmed: 21444623
J Clin Epidemiol. 2008 Jan;61(1):76-86
pubmed: 18083464
Stat Med. 2017 Dec 10;36(28):4529-4539
pubmed: 27891652
South Med J. 2016 Jun;109(6):365-9
pubmed: 27255094
J Am Coll Cardiol. 2002 Feb 6;39(3):471-80
pubmed: 11823086
J Am Med Inform Assoc. 2012 Mar-Apr;19(2):263-74
pubmed: 21984587
Am J Gastroenterol. 2012 Feb;107(2):247-52
pubmed: 21931378
BMC Med Res Methodol. 2014 Dec 22;14:137
pubmed: 25532820
Hepatology. 2016 Aug;64(2):569-81
pubmed: 26991920
J Am Coll Cardiol. 2013 Jan 29;61(4):391-403
pubmed: 23219302
Gastroenterology. 2001 Jan;120(1):170-8
pubmed: 11208726
Hepatology. 2015 Aug;62(2):584-90
pubmed: 25846824
Hepatogastroenterology. 2014 Jul-Aug;61(133):1170-4
pubmed: 25436277
Intensive Care Med. 2012 Jan;38(1):40-6
pubmed: 22042520
JAMA Intern Med. 2014 Jul;174(7):1095-107
pubmed: 24820131
Gastroenterology. 2015 Dec;149(7):1731-1741.e3
pubmed: 26327134
JAMA. 2011 Oct 19;306(15):1688-98
pubmed: 22009101
Pharmacoepidemiol Drug Saf. 2011 Jul;20(7):689-99
pubmed: 21626605
Arch Intern Med. 2000 Apr 24;160(8):1074-81
pubmed: 10789599
Ann Intern Med. 2011 Oct 18;155(8):520-8
pubmed: 22007045
J Am Med Inform Assoc. 2017 Nov 1;24(6):1052-1061
pubmed: 28379439
JAMA. 2013 Jan 23;309(4):381-91
pubmed: 23340640
Hepatology. 2016 Jul;64(1):200-8
pubmed: 26690389
J Clin Epidemiol. 1996 Dec;49(12):1373-9
pubmed: 8970487
Stat Med. 2004 Aug 30;23(16):2567-86
pubmed: 15287085
Eur J Cardiothorac Surg. 2013 Jan;43(1):206
pubmed: 22728228
PLoS One. 2011 Feb 23;6(2):e16110
pubmed: 21373178