Background Current guideline recommend an early invasive strategy for NSTEMI patients (Class IIA)....

1
Background Current guideline recommend an early invasive strategy for NSTEMI patients (Class IIA). However, 67% US hospitals have no catheterization capability. Patients transferred for invasive strategy are usually younger with less comorbidities compared with non-transferred patients. Thus, hospitals with high proportion of transfer patients should have better outcomes of care. However, the association of transfer proportion and outcomes in transfer-out hospitals has not been previously examined. The association of transfer-out rates from hospitals without revascularization capabilities and mortality risk among older NSTEMI patients Lan Shen, MD, Shuang Li, MS, Laine Thomas, Ph.D, Bimal R. Shah, MD, MBA, Tracy Y. Wang, MD, MHS, Karen Alexander, MD, Eric D. Peterson, MD, MPH, He Ben, MD, Ph.D, Matthew T. Roe, MD, MHS Duke Clinical Research Institute, Durham, NC, U.S.A; Shanghai Renji Hospital, Cardiology department, Shanghai, China Methods 5,678 eligible NSTEMI patients in 65 hospitals without PCI / CABG capabilities were identified in the CRUSADE registry from 2003 to 2006, were linked to Medicare claims data to assess longitudinal outcomes. The distribution of transfer-out rate among all hospitals was examined. Based on the distribution rate, high transfer-out hospitals were defined as having >40% of all eligible patients transferred out, whereas low transfer-out hospitals were defined as having ≤40% of all eligible patients. Baseline, presentation features, in-hospital procedures and discharge medications were described by transfer rate status. Wilcoxon- rank rum test was used for comparing continuous variables and chi-square test was used for categorical variables. Overall baseline risk profiles were compared using the CRUSADE long-term mortality risk score between patients in the 2 groups of hospitals. Multivariable Cox proportional hazard model was used to assess the association between the proportion of transfer out and 30-day, 6-month and 3-year mortality. Results . Conclusions Among older patients >65 years old, hospitals with high transfer-out proportions have more low risk patients, with more aggressive acute medication treatment. Hospitals with higher transfer-out proportions have lower observed mortality rate of short-term and long-term follow-up. However, such survival advantage disappears after adjustment for baseline characteristics. The difference in hospital-level case mix can explain the lack of difference in the adjusted long-term mortality risk between hospital categories. Our study support that older patients who are admitted in non-PCI hospitals should undergo early invasive management. Although older patients are usually excluded from clinical trials, no information guide decision making for older patients. Our exploratory study Acknowledgments No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the manuscript, and its final contents. Contact Lan Shen, MD, MS. Shanghai Renji Hospital, Duke Clinical Research Institute Tel: 919 -641-9233 Fax: 919-668-7061 Email: [email protected] Variables Overall Patients=56 78 Hospitals=6 5 Low transfer Patients=27 15 Hospitals High transfer Patients=29 63 Hospitals=3 Age (Median, IQR) 79 (73, 85) 80 (73, 86) 79 (73,85) Male sex (%) 48 46 50 Medical history (%) Smoking 11 11 11 Hypertensio n 76 76 76 Diabetes mellitus 36 35 37 Renal insufficien cy 19 21 18 Dyslipidemi a 48 44 51 Prior MI 29 27 31 Prior PCI 15 13 17 Prior CABG 21 22 20 Prior CHF 27 30 24 Presentation characteristics (%) Table 2: Impact of transfer-out rate on outcomes CRUSADE registry CRUSADE was a national quality improvement initiative designed to promote evidence-based treatment of hospitalized patients with non–ST-segment elevation ACS. Patient data were collected retrospectively via chart review from July, 2001 through December, 2006. More than 500 hospitals in the US Limitation Given the retrospective nature of the study, unmeasured confounders cannot be excluded, especially comorbidities which prevent patients from being transferred. The small number of hospitals in our study did not include all non-revascularization hospitals in U.S, thus it is not representative for hospitals outside of CRUSADE. Our study serves as an exploratory study. CRUSADE did not collect long-term medication use, so the impact cannot be measured. Figure 1: Distribution of transfer-out rate among all hospitals. Figure 2. Percentage of patients in different quartiles of baseline CRUSADE risk score between 2 groups of hospitals. Table 1. Hospital characteristics Variables Overall (n=65) Low transfer Hospitals (n= 27) High transfer Hospitals (n= 38) Region (%) West 11 11 11 Northeast 35 37 34 Midwest 11 19 5 South 43 33 50 Type of hospital (%) No service 40 30 47 Cath Lab Only 60 70 53 Teaching Hospital (%) Non-Academic 94 89 97 Academic 6 11 3 Total hospital 202 230 167 Table 3. Impact of transfer out rate on outcomes (high transfer-out vs. low transfer- out). Outcomes Observed mortality rate Unadjusted HR* (95% CI) Adjusted HR* (95% CI) Mortality High vs. Low 30-day mortality 10% vs. 14% 0.75 (0.59- 0.96) 0.92 (0.77- 1.10) 6-month mortality 22% vs. 27% 0.80 (0.65- 0.97) 0.95 (0.83- 1.08) 3-year mortality 46% vs. 52% 0.84 (0.73- 0.96) 0.99 (0.89- 1.09) *Hazard ratio of the outcomes between high transfer-out rate hospitals vs. low transfer- out hospitals (reference group) CI, confidence interval; HR, hazard ratio Adjusted for: age, male sex, race, weight, dyslipidemia, initial HCT with knot at 35%, initial troponin ratio with two knots (with knots at 5 and 50), prior stroke, diabetes mellitus, signs of heart failure, initial serum creatinine, initial systolic blood pressure, initial heart rate, prior percutaneous coronary intervention (PCI), electrocardiographic changes (ST depression, transient ST elevation, both [vs. neither]), hypertension, prior coronary artery bypass graft (CABG) , PCI, CABG procedures

Transcript of Background Current guideline recommend an early invasive strategy for NSTEMI patients (Class IIA)....

Page 1: Background Current guideline recommend an early invasive strategy for NSTEMI patients (Class IIA). However, 67% US hospitals have no catheterization capability.

Background• Current guideline recommend an early invasive

strategy for NSTEMI patients (Class IIA). However, 67% US hospitals have no catheterization capability.

• Patients transferred for invasive strategy are usually younger with less comorbidities compared with non-transferred patients.

• Thus, hospitals with high proportion of transfer patients should have better outcomes of care.

• However, the association of transfer proportion and outcomes in transfer-out hospitals has not been previously examined.

The association of transfer-out rates from hospitals without revascularization capabilities and mortality risk among older NSTEMI patientsLan Shen, MD, Shuang Li, MS, Laine Thomas, Ph.D, Bimal R. Shah, MD, MBA, Tracy Y. Wang, MD, MHS, Karen Alexander, MD, Eric D. Peterson, MD, MPH, He Ben, MD, Ph.D, Matthew T. Roe, MD, MHS

Duke Clinical Research Institute, Durham, NC, U.S.A; Shanghai Renji Hospital, Cardiology department, Shanghai, China

Methods• 5,678 eligible NSTEMI patients in 65 hospitals

without PCI / CABG capabilities were identified in the CRUSADE registry from 2003 to 2006, were linked to Medicare claims data to assess longitudinal outcomes.

• The distribution of transfer-out rate among all hospitals was examined. Based on the distribution rate, high transfer-out hospitals were defined as having >40% of all eligible patients transferred out, whereas low transfer-out hospitals were defined as having ≤40% of all eligible patients.

• Baseline, presentation features, in-hospital procedures and discharge medications were described by transfer rate status. Wilcoxon-rank rum test was used for comparing continuous variables and chi-square test was used for categorical variables.

• Overall baseline risk profiles were compared using the CRUSADE long-term mortality risk score between patients in the 2 groups of hospitals.

• Multivariable Cox proportional hazard model was used to assess the association between the proportion of transfer out and 30-day, 6-month and 3-year mortality.

Results

.

Conclusions

• Among older patients >65 years old, hospitals with high transfer-out proportions have more low risk patients, with more aggressive acute medication treatment.

• Hospitals with higher transfer-out proportions have lower observed mortality rate of short-term and long-term follow-up. However, such survival advantage disappears after adjustment for baseline characteristics.

• The difference in hospital-level case mix can explain the lack of difference in the adjusted long-term mortality risk between hospital categories.

• Our study support that older patients who are admitted in non-PCI hospitals should undergo early invasive management. Although older patients are usually excluded from clinical trials, no information guide decision making for older patients. Our exploratory study supported older patients to be recommended to aggressive treatment.

Acknowledgments

No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the manuscript, and its final contents.

Contact

Lan Shen, MD, MS.

Shanghai Renji Hospital,

Duke Clinical Research Institute

Tel: 919 -641-9233

Fax: 919-668-7061

Email: [email protected]

Variables

Overall

Patients=5

678

Hospitals=

65

Low

transfer

Patients=2

715

Hospitals

=27

High

transfer

Patients=2

963

Hospitals=

38

Age(Median, IQR)

79 (73, 85) 80 (73, 86) 79 (73,85)

Male sex (%) 48 46 50

Medical history (%)

Smoking 11 11 11

Hypertensio

n

76 76 76

Diabetes mellitus 36 35 37

Renal insufficiency 19 21 18

Dyslipidemia48 44 51

Prior MI 29 27 31

Prior PCI 15 13 17

Prior

CABG21 22 20

Prior CHF 27 30 24

Presentation characteristics (%)

Signs of

CHF36 38 35

Heart rate 88 (74, 106) 89 (75, 106) 87 (73,105)

SBP,

mmHg

143 (120,

164)

141 (119,

162)

145 (122,

166) ST depression 24 22 25

CRUSADE Long-term Mortality Risk Score

36

(25, 47)

37

(26, 48)

34

(24, 46)

Medications within 24 hours (%)

Aspirin 92 91 94 Anticoagula-tion

77 74 79

Clopidogrel 43 44 43

GP IIb/IIIa 28 21 34Diagnostic Cath(%) 31 43 21

Table 2: Impact of transfer-out rate on outcomes

CRUSADE registry

CRUSADE was a national quality improvement initiative designed to promote evidence-based treatment of hospitalized patients with non–ST-segment elevation ACS. Patient data were collected retrospectively via chart review from July, 2001 through December, 2006. More than 500 hospitals in the US participated including more than 200,000 patients.

Limitation

• Given the retrospective nature of the study, unmeasured confounders cannot be excluded, especially comorbidities which prevent patients from being transferred.

• The small number of hospitals in our study did not include all non-revascularization hospitals in U.S, thus it is not representative for hospitals outside of CRUSADE. Our study serves as an exploratory study.

• CRUSADE did not collect long-term medication use, so the impact cannot be measured.

Figure 1: Distribution of transfer-out rate among all hospitals. Figure 2. Percentage of patients in different quartiles of baseline CRUSADE risk score between 2 groups of hospitals.

Table 1. Hospital characteristics

Variables

Overall

(n=65)

Low transfer

Hospitals (n= 27)

High transfer

Hospitals(n= 38)

Region (%)      

West 11 11 11

Northeast 35 37 34

Midwest 11 19 5

South 43 33 50

Type of hospital

(%)

     

No service 40 30 47

Cath Lab Only 60 70 53

Teaching Hospital

(%)

     

Non-Academic 94 89 97

Academic 6 11 3

Total hospital Beds (median, IQR)

202 (131, 278)

230 (171,300)

167 (110, 260)

Table 3. Impact of transfer out rate on outcomes (high transfer-out vs. low transfer-out).

Outcomes

Observed

mortality

rate

Unadjusted

HR* (95%

CI)

Adjusted

HR* (95%

CI)

MortalityHigh vs. Low

30-day mortality10% vs.

14%0.75

(0.59-0.96)0.92

(0.77-1.10)

6-month mortality

22% vs. 27%

0.80 (0.65-0.97)

0.95 (0.83-1.08)

3-year mortality

46% vs. 52%

0.84(0.73-0.96)

0.99 (0.89-1.09)

*Hazard ratio of the outcomes between high transfer-out

rate hospitals vs. low transfer-out hospitals (reference

group)

CI, confidence interval; HR, hazard ratioAdjusted for: age, male sex, race, weight, dyslipidemia, initial HCT with knot at 35%, initial troponin ratio with two knots (with knots at 5 and 50), prior stroke, diabetes mellitus, signs of heart failure, initial serum creatinine, initial systolic blood pressure, initial heart rate, prior percutaneous coronary intervention (PCI), electrocardiographic changes (ST depression, transient ST elevation, both [vs. neither]), hypertension, prior coronary artery bypass graft (CABG) , PCI, CABG procedures used within 7 days post transfer out.