RATIONAL, AIMS AND OBJECTIVES: The study aims to determine the extent to which the addition of post-admission information via time-dependent covariates improved the ability of a survival model to predict the daily risk of hospital death.
Using administrative and laboratory data from adult inpatient hospitalizations at our institution between 1 April 2004 and 31 March 2009, we fit both a time-dependent and a time-fixed Cox model for hospital mortality on a randomly chosen 66% of hospitalizations. We compared the predictive performance of these models on the remaining hospitalizations.
All comparative measures clearly indicated that the addition of time-dependent covariates improved model discrimination and prominently improved model calibration. The time-dependent model had a significantly higher concordance probability (0.879 versus 0.811) and predicted significantly closer to the number of observed deaths within all risk deciles. Over the first 32 admission days, the integrated discrimination improvement (IDI) and net reclassification improvement (NRI) were consistently above zero (average IDI of +0.0200 and average NRI of 62.7% over the first 32 days).
The addition of time-dependent covariates significantly improved the ability of a survival model to predict a patient's daily risk of hospital death. Researchers should consider adding time-dependent covariates when seeking to improve the performance of survival models.
Adverse events (AEs) are adverse outcomes caused by medical care. Several studies have indicated that a substantial number of patients experience AEs before or during hospitalization. However, few data describe AEs after hospital discharge. We determined the incidence, severity, preventability and ameliorability of AEs in patients discharged from the general internal medicine service of a Canadian hospital.
At a multisite Canadian teaching hospital, we prospectively studied patients who were consecutively discharged home or to a seniors' residence from the general internal medicine service during a 14-week interval in 2002. We used telephone interview and chart review to identify outcomes after discharge. Two physicians independently reviewed each outcome to determine if the patient experienced an AE. The severity, preventability and ameliorability of all AEs were classified.
During the study period, outcomes were determined for 328 of the 361 eligible patients, who averaged 71 years of age (interquartile range 54-81 years). After discharge, 76 of the 328 patients experienced at least 1 AE (overall incidence 23%, 95% confidence interval [CI] 19%-28%). The AE severity ranged from symptoms only (68% of the AEs) or symptoms associated with a nonpermanent disability (25%) to permanent disability (3%) or death (3%). The most common AEs were adverse drug events (72%), therapeutic errors (16%) and nosocomial infections (11%). Of the 76 patients, 38 had an AE that was either preventable or ameliorable (overall incidence 12%, 95% CI 9%-16%).
Approximately one-quarter of patients in our study had an AE after hospital discharge, and half of the AEs were preventable or ameliorable.
Cites: JAMA. 1991 Apr 17;265(15):1977-812008027
Cites: JAMA. 1991 Apr 17;265(15):1957-602008024
Cites: Med J Aust. 1995 Nov 6;163(9):458-717476634
Many studies demonstrate a high rate of treatment-related adverse outcomes or adverse events. No studies have prospectively evaluated adverse events in patients discharged home from the emergency department (ED).
To describe the types of adverse events in patients discharged home from an ED.
PATIENTS who were sent home directly from the ED of an urban, academic teaching hospital in Ottawa, Canada.
Patient records were reviewed to identify demographic and medical history information. Two weeks following the ED visit, patients completed a standard telephone interview to record post ED visit outcomes. Two physicians reviewed outcomes to identify all adverse events and their cause.
Follow-up was complete for 399 of 408 enrolled patients. The median age was 49 years (interquartile range 36-68) and 50% were male. The most common diagnosis was "chest pain", occurring in 74 patients (18%), followed by "bone and joint disorders" in 55 patients (14%). 24 patients experienced an adverse event (incidence 6% (95% CI 4% to 9%)), of which 17 were preventable (incidence 4% (95% CI 3% to 7%)). Five of the unpreventable adverse events were medication side effects and two were minor, procedure-related complications. Of all 24 adverse events, 15 (63%; 95% CI 43 to 79%) led to an additional ED visit or a hospitalisation. Preventable adverse events occurred in 5 of 78 chest pain patients (incidence 6% (95% CI 3% to 14%)).
Most adverse events occurring following an ED visit are preventable and often relate to diagnostic or management errors.
Cites: Ann Emerg Med. 1999 Aug;34(2):155-910424915
Cites: Ann Intern Med. 1999 Dec 21;131(12):909-1810610641
It was the objective of this study to determine the proportion of patients who undergo an appropriate diagnostic work-up following a D-dimer test performed to evaluate suspected pulmonary embolism (PE) or deep vein thrombosis (DVT). We performed a retrospective cohort study at a tertiary care hospital. We included patients if they underwent D-dimer testing between 2002 and 2005, if the D-dimer was performed for evaluation of VTE, and if the D-dimer test was successful. We classified: the patients' clinical probability of DVT or PE according to the Wells models, the imaging results, and the appropriateness of the testing algorithm. Of 1,000 randomly selected patients, 863 met our study criteria. Seven hundred nineteen patients (83%) had testing during an emergency department visit, while 144 were tested as inpatients (17%). Physicians performed the D-dimer test to evaluate DVT and PE in 238 (28%) and 625 (72%) patients, respectively. Overall, the testing strategy was appropriate in 69% (95% confidence interval [CI]: 66%-72%) of cases. The testing strategy was more likely to be appropriate for emergency department versus inpatients (75% vs. 39%, p
To assist physicians with difficult decisions about hospital admission for patients with acute exacerbation of chronic obstructive pulmonary disease (COPD) presenting in the emergency department, we sought to identify clinical characteristics associated with serious adverse events.
We conducted this prospective cohort study in 6 large Canadian academic emergency departments. Patients were assessed for standardized clinical variables and then followed for serious adverse events, defined as death, intubation, admission to a monitored unit or new visit to the emergency department requiring admission.
We enrolled 945 patients, of whom 354 (37.5%) were admitted to hospital. Of 74 (7.8%) patients with a subsequent serious adverse event, 36 (49%) had not been admitted after the initial emergency visit. Multivariable modelling identified 5 variables that were independently associated with adverse events: prior intubation, initial heart rate = 110/minute, being too ill to do a walk test, hemoglobin
Cites: Chest. 1991 Mar;99(3):595-91995214
Cites: BMJ. 1995 Sep 2;311(7005):594-77663253
Cites: Am J Respir Crit Care Med. 1998 May;157(5 Pt 1):1418-229603117
Cites: Ann Intern Med. 2006 Feb 7;144(3):165-7116461960
Between 5% and 10% of patients die or are urgently readmitted within 30 days of discharge from hospital. Readmission risk indexes have either excluded acute diagnoses or modelled them as multiple distinct variables. In this study, we derived and validated a score summarizing the influence of acute hospital diagnoses and procedures on death or urgent readmission within 30 days.
From population-based hospital abstracts in Ontario, we randomly sampled 200 000 discharges between April 2003 and March 2009 and determined who had been readmitted urgently or died within 30 days of discharge. We used generalized estimating equation modelling, with a sample of 100 000 patients, to measure the adjusted association of various case-mix groups (CMGs-homogenous groups of acute care inpatients with similar clinical and resource-utilization characteristics) with 30-day death or urgent readmission. This final model was transformed into a scoring system that was validated in the remaining 100 000 patients.
Patients in the derivation set belonged to 1 of 506 CMGs and had a 6.8% risk of 30-day death or urgent readmission. Forty-seven CMG codes (more than half of which were directly related to chronic diseases) were independently associated with this outcome, which led to a CMG score that ranged from -6 to 7 points. The CMG score was significantly associated with 30-day death or urgent readmission (unadjusted odds ratio for a 1-point increase in CMG score 1.52, 95% confidence interval [CI] 1.49-1.56). Alone, the CMG score was only moderately discriminative (C statistic 0.650, 95% CI 0.644-0.656). However, when the CMG score was added to a validated risk index for death or readmission, the C statistic increased to 0.759 (95% CI 0.753-0.765). The CMG score was well calibrated for 30-day death or readmission.
In this study, we developed a scoring system for acute hospital diagnoses and procedures that could be used as part of a risk-adjustment methodology for analyses of postdischarge outcomes.
As electronic patient data from automated hospital databases become increasingly available, it is important to explore the ways in which these data could be used for the purposes other than patient care, such as quality assurance and improvement.
To determine if information from automated patient databases can be used to derive a model that can predict patients' daily risk of death in hospital. Such a model could be used to improve the ability to risk-adjust hospital mortality rates.
Retrospective cohort study of 159,794 hospitalizations at The Ottawa Hospital between April 1, 2004 and March 31, 2009. The model was derived using time-dependent Cox regression methods on a random two-thirds of admissions. The model was validated by applying the coefficients to the other third of admissions.
Inpatient mortality was 5%. The final model included: patient age; admission type; intensive care unit status; alternative level of care status; and separate scores for patient comorbidity, in-hospital procedures, and acute illness (using information from 14 laboratory tests). In the validation set, the model had excellent discrimination (c-statistic 0.879, 95% confidence interval: 0.872-0.886) and calibration in all risk strata over all admission days.
We found that information from our hospital's automated patient databases could be used to accurately predict patients' daily risk of death in hospital. The predictions from this model could be used in quality of care analyses to more accurately risk-adjust hospital mortality rates and by hospitals to improve triage processes and patient flow.
Clinicians informally assess changes in patients' status over time to prognosticate their outcomes. The incorporation of trends in patient status into regression models could improve their ability to predict outcomes. In this study, we used a unique approach to measure trends in patient hospital death risk and determined whether the incorporation of these trend measures into a survival model improved the accuracy of its risk predictions.
We included all adult inpatient hospitalizations between 1 April 2004 and 31 March 2009 at our institution. We used the daily mortality risk scores from an existing time-dependent survival model to create five trend indicators: absolute and relative percent change in the risk score from the previous day; absolute and relative percent change in the risk score from the start of the trend; and number of days with a trend in the risk score. In the derivation set, we determined which trend indicators were associated with time to death in hospital, independent of the existing covariates. In the validation set, we compared the predictive performance of the existing model with and without the trend indicators.
Three trend indicators were independently associated with time to hospital mortality: the absolute change in the risk score from the previous day; the absolute change in the risk score from the start of the trend; and the number of consecutive days with a trend in the risk score. However, adding these trend indicators to the existing model resulted in only small improvements in model discrimination and calibration.
We produced several indicators of trend in patient risk that were significantly associated with time to hospital death independent of the model used to create them. In other survival models, our approach of incorporating risk trends could be explored to improve their performance without the collection of additional data.
Cites: Stat Med. 2008 Jan 30;27(2):157-72; discussion 207-1217569110
Cites: Med Care. 2008 Mar;46(3):232-918388836
Cites: J Am Coll Cardiol. 2008 Sep 16;52(12):997-100318786480
Cites: Med Care. 2009 Jun;47(6):626-3319433995
Cites: Qual Life Res. 2006 Oct;15(8):1297-30616830258
Cites: Stat Med. 2011 Jan 15;30(1):11-2121204120
Cites: Eur Heart J. 2000 Jan;21(2):146-5310637088
Cites: J Am Soc Nephrol. 2004 Jun;15(6):1597-60515153571
Cites: Lifetime Data Anal. 1998;4(2):109-209658770
Cites: Qual Life Res. 2009 Oct;18(8):1019-2719697155
Several randomized trials have found that discharge planning improves outcomes for hospitalized patients. We do not know if adding a clinical nurse specialist (CNS) to physician teams in hospitals that already have discharge planning services makes a difference.
In 2 teaching hospitals, patients were randomly assigned to regular hospital care or care with a clinical nurse specialist. The clinical nurse specialist facilitated hospital care by retrieving preadmission information, arranging in-hospital consultations and investigations, organizing postdischarge follow-up visits, and checking up on patients postdischarge with a telephone call. In-hospital outcomes included mortality and length of stay. Postdischarge outcomes included time to readmission or death, patient satisfaction, and the risk of adverse event. Adverse events were poor outcomes due to medical care rather than the natural history of disease.
A total of 620 sequential patients were randomized (CNS n = 307, control n = 313), of which 361 were followed after discharge from hospital (CNS n = 175, control n = 186). The groups were similar for the probability of in-hospital death (CNS 9.3% vs control 9.7%) or being discharged to the community (58.0% vs 60.0%). The groups did not differ for postdischarge outcomes including readmission or death (21.6% vs 15.6%; P = 0.16) or risk of adverse event (23.6% vs 22.8%). Mean [SD] patient ratings of overall quality of care on a scale of 10 was higher in the clinical nurse specialist group (8.2 [2.2] vs 7.6 [2.4]; P = 0.052).
The addition of a clinical nurse specialist to a medical team improved patient satisfaction but did not impact hospital efficiency or patient safety.
The effects of hospital-acquired Clostridium difficile infection (CDI) on patient outcomes are incompletely understood. We conducted this study to determine the independent impact of hospital-acquired CDI on in-hospital mortality after adjusting for the time-varying nature of CDI and baseline mortality risk at hospital admission.
This retrospective observational study used data from the Ottawa Hospital (Ottawa, Ontario, Canada) data warehouse. Inpatient admissions with a start date after July 1, 2002, and a discharge date before March 31, 2009, were included. Stratified analyses and a Cox multivariate proportional hazards regression model were used to determine if hospital-acquired CDI was associated with time to in-hospital death.
A total of 136 877 admissions were included. Hospital-acquired CDI was identified in 1393 admissions (overall risk per admission, 1.02%; 95% confidence interval [CI], 0.97%-1.06%). The risk of hospital-acquired CDI significantly increased as the baseline mortality risk increased: from 0.2% to 2.6% in the lowest to highest deciles of baseline risk. Hospital-acquired CDI significantly increased the absolute risk of in-hospital death across all deciles of baseline risk (pooled absolute increase, 11%; 95% CI, 9%-13%). Cox regression analysis revealed an average 3-fold increase in the hazard of death associated with hospital-acquired CDI (95% CI, 2.4-3.7); this hazard ratio decreased with increasing baseline mortality risk.
Hospital-acquired CDI was independently associated with an increased risk of in-hospital death. Across all baseline risk strata, for every 10 patients acquiring the infection, 1 person died.