SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Background

Electronic health record systems (EHR) are expected to facilitate higher quality patient care; however, studies evaluating EHR effectiveness in improving care have yielded mixed results.

Hypothesis

Implementation of a performance improvement system in outpatient practices with EHR may better demonstrate the value of EHR in improving quality.

Methods

The Registry to Improve the Use of Evidence-Based Heart Failure Therapies in the Outpatient Setting (IMPROVE HF) prospectively evaluated the effectiveness of a performance improvement initiative on use of evidence-based therapies for patients with heart failure (HF) or prior MI and LVSD. This study assessed improvement in the use of 7 quality measures from baseline to 24 months.

Results

Complete data were available for 155 of 167 (92.8%) practices; 78 (50.3%) used EHR always, 15 (9.7%) switched to EHR, and 61 (39.4%) used paper always. EHR-always practices had significantly improved adherence to 5 measures at 24 months, and EHR-switched or paper-always practices had improved adherence to 6 measures. With a single exception, there were no significant differences in the magnitude of improvements in use of guideline-recommended care among the 3 practice types. Performance on individual quality measures was also similar at 24 months.

Conclusions

Implementation of the performance improvement intervention enhanced use of guideline-recommended HF therapies among outpatient cardiology practices. However, practices using or converting to EHR did not achieve greater improvements in quality of HF care than practices using paper systems. These findings raise doubts about whether implementation of EHR nationally will translate into better outpatient quality of care. Copyright © 2010 Wiley Periodicals, Inc.

IMPROVE HF is registered at http://www.clinicaltrials. gov, study number NCT00303979, and is supported by Medtronic, Inc, Minneapolis, Minnesota.

Complete disclosures may be found in the online version of the article.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Electronic health record systems (EHRs) are expected to improve adherence to evidence-based guidelines and are widely thought to promote higher-quality, more-efficient patient care.1,2 The Institute of Medicine has strongly advocated EHR use to improve healthcare quality, and the Health Information Technology for Economic and Clinical Health (HITECH) provisions of the American Reinvestment and Recovery Act of 2009 are providing for large financial incentives and penalties to encourage adoption of EHR.3,4 However, studies evaluating the effectiveness of EHR in improving the quality of patient care have yielded mixed results, particularly in the outpatient setting.5–10 Some prior studies have demonstrated better quality as a result of EHR implementation.10 However, others using national data found no difference in quality of care between ambulatory care provided with and without EHR.6,11,12 A baseline analysis of data from the Registry to Improve the Use of Evidence-Based Heart Failure Therapies in the Outpatient Setting (IMPROVE HF) demonstrated that use of an EHR was associated with no or only modest differences in the quality of care for patients with heart failure (HF) in U.S. outpatient cardiology and multispecialty practices.13

A lack of association between EHR and outpatient quality of care may reflect variable deployment of clinical decision support functionality within an EHR and the fact that there have, in general, not been performance-improvement initiatives coupled with EHR deployment. Implementation of a performance-improvement system in outpatient practices with EHR may allow for greater gains in quality of care and better demonstrate the value of the systems in improving quality. IMPROVE HF tested the impact of a practice-based performance-improvement system on the use of guideline-recommended HF therapies during a 24-month period. IMPROVE HF provides an important opportunity to determine whether a performance-improvement intervention would lead to greater use of evidence-based, guideline-recommended HF care among practices utilizing EHR.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

IMPROVE HF was a prospective study designed to evaluate the effectiveness of a practice-specific performance-improvement intervention on the use of evidence-based therapies for patients with HF or prior myocardial infarction (MI) and left ventricular systolic dysfunction (LVSD) in outpatient cardiology practices. The methods have been described previously in detail.14,15 Community and academic single- or multispecialty cardiology practices throughout the United States were invited to participate. Patients with a clinical diagnosis of HF or post-MI and LVSD were eligible for enrollment. LVSD was determined quantitatively by a left ventricular ejection fraction ≤35% on the most recent echocardiogram, nuclear multiple gated acquisition scan, contrast ventriculogram, or magnetic resonance imaging scan or qualitatively based on findings of moderate to severe LVSD.14,15 Patients with a noncardiovascular medical condition associated with an estimated survival of ≤1 year and those who had undergone cardiac transplantation were excluded.

After the medical records of potentially eligible patients were screened, an average of 90 patients from each practice were randomly selected for inclusion in the study. Thirty-four specially trained, centralized abstractors reviewed the medical charts of all eligible patients to determine therapy utilization rates at baseline and at prespecified time points throughout a 24-month period after implementation of the performance-improvement intervention. Demographic and clinical characteristics, New York Heart Association (NYHA) functional class, QRS duration, laboratory results, diagnostic tests, treatments, and the provision of HF education were recorded, as were documented contraindications and other reasons for not administering the selected therapies to eligible patients. Multiple measures were established to ensure the accuracy of data collection, including ongoing training by the IMPROVE HF Steering Committee, automated data quality checks, monthly data quality reports, and periodic auditing of collected data in comparison with source documentation.14,15 Average interrater reliability was 0.82 (κ statistic); the mean concordance rate for the audits was 94.5% (range, 92.3%–96.3%).

Practice Characteristics

Practice characteristics were collected by survey. Information collected included practice type and geographic region; specialization (single- or multispecialty); number of patients managed annually; number of cardiologists, electrophysiologists, interventionalists, and advanced practice nurses on staff; presence of a dedicated HF clinic; and type of medical record system (paper only, EHR only, or paper + EHR). Practices that reported use of EHR were asked to provide additional information about the type of system used (brand, version) and the capability to customize the screens and records.13 Information about the current type of medical record system was confirmed by study personnel during the baseline and 24-month data collections. Practices were required to obtain institutional review board approval or waivers to participate in the IMPROVE HF registry. Outcome Sciences, Inc (Cambridge, MA), served as the registry coordinating center.

Evidence-Based Therapies

Use of 7 evidence-based HF therapies was evaluated in IMPROVE HF: (1) angiotensin-converting enzyme inhibitor/angiotensin receptor blocker (ACEI/ARB), (2) β-blocker, (3) aldosterone antagonist, (4) anticoagulant therapy for atrial fibrillation or flutter, (5) cardiac resynchronization therapy (CRT) with a defibrillator (CRT-D) or pacemaker (CRT-P), (6) implantable cardioverter-defibrillator (ICD or CRT-D), and (7) HF education.14,15 Patients were considered eligible for a therapy if they met the indications for that therapy and had no documented contraindications, intolerance, or other reasons for not receiving it. Documentation of NYHA functional class was required to be considered eligible for an ICD, CRT, or aldosterone antagonist. Four of the therapies (ACEI or ARB, β-blocker, anticoagulation for atrial fibrillation, and HF education) are current American College of Cardiology/American Heart Association (ACC/AHA) HF outpatient performance measures and are endorsed by the National Quality Forum.16

Performance-Improvement Intervention

The IMPROVE HF performance-improvement intervention was designed to facilitate complex knowledge transfer and improve systems for treating patients with HF.14 The investigators and practice support personnel attended a 1-day workshop after completion of baseline data collection. Study goals, guideline recommendations, the IMPROVE HF tool kit, performance-improvement methodology, incentives to promote change in clinical practice, tips on how to serve as champions and engage colleagues in the performance intervention, and strategies to effectively use collected data to provide practice-specific performance feedback were provided. The intervention included a guideline-based clinical decision support tool kit, educational materials, practice-specific data reports, benchmarked quality-of-care reports, and structured educational and collaborative opportunities. As part of an enhanced treatment plan, IMPROVE HF provided evidence-based best practices algorithms, clinical pathways, standardized encounter forms, checklists, pocket cards, chart stickers, and patient education and other materials to facilitate improved management of outpatients with HF. Use of the tools was highly encouraged, and practices with EHR were advised to integrate these tools into the EHR clinical decision support functionality. Practices could adopt or modify tools at their discretion. A web-based system provided quality-of-care reports for each practice that included benchmark comparisons with regional and national cardiology practices. Participating practices were encouraged to participate in bimonthly educational and collaborative web-based seminars and to continually evaluate, refine, and reassess care delivery throughout the intervention phase of the study.14

Statistical Analyses

The cohort for these analyses comprised patients with data available at both baseline and the 24-month follow-up assessment. Descriptive patient and practice characteristics were summarized according to 3 EHR groups: (1) paper-always (ie, practices using only paper records at both baseline and 24 months), (2) EHR-always (ie, EHR or mixed at baseline and 24 months), and (3) EHR-switched (ie, paper only at baseline and EHR only or mixed at 24 months) groups. P values were determined by analysis of variance (ANOVA) for continuous variables and by χ2 analysis for categoric values. P < 0.05 was considered statistically significant.

A univariate generalized estimating equation (GEE) model was used to determine the improvement in each quality measure in the 3 groups. Improvement was defined as treatment at 24 months of patients eligible for each quality measure at both time points but not treated at baseline. Odds ratios and 95% confidence intervals were calculated to compare improvement between the groups (EHR-always vs paper-always and EHR-switched vs paper-always). Characteristics with P ≤ 0.10 were included in a multivariate GEE model to determine whether use of EHR was associated with improvement in each of the quality measures. P < 0.05 was considered statistically significant. Univariate and multivariate GEE models were also used to determine whether use of EHR was associated with conformity (adherence) to the 7 quality measures at 24 months. Odds ratios and 95% confidence intervals were calculated and compared between the 3 groups. The multivariate GEE analyses included characteristics with P ≤ 0.10 as covariates, and P < 0.05 was considered statistically significant. Analysis of covariance was used to compare absolute improvement and adherence at 24 months between the 3 groups at the practice level. Covariates included the baseline treatment rate as well as patient and practice characteristics with P < 0.10 in the ANOVA model. P < 0.05 was considered statistically significant. Statistical analyses were performed by 1 of the authors (Y.L.), an employee of Medtronic, under the direction of the IMPROVE HF Steering Committee. Analyses were completed using SAS statistical software, version 9.1 (SAS Institute, Cary, NC).

IMPROVE HF is registered at http://www.clinicaltrials.gov, study number NCT00303979. The design and conduct of the IMPROVE HF registry and investigations are in accordance with the Declaration of Helsinki.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

At baseline, 167 outpatient cardiology practices contributing data on 15 177 patients were enrolled in IMPROVE HF. Of these, 155 (92.8%) practices with 7893 patients had complete data available at 24 months. Seventy-eight of the 155 practices (50.3%) reported using EHR-always, 15 (9.7%) EHR-switched, and 61 (39.4%) paper-always. One practice switched from an EHR to a paper system during the course of the study and was excluded from the analyses. The number of patients in each study group was as follows: EHR-always, 4220; EHR-switched, 723; and paper-always, 2950.

Practice characteristics of the 3 groups are shown in Table 1. There were some geographic differences in the practice types stratified by EHR, with a greater proportion of practices in the EHR-always group in the West and more practices in the Northeast and Central regions switching to EHR. The proportion of practices in the paper-always group was highest in nonuniversity and nonteaching facilities. Other practice characteristics are listed in Table 1.

Table 1. Baseline Practice Characteristics by Type of Medical Record System
CharacteristicEHR-Always, n = 4220aEHR-Switched, n = 723aPaper-Always, n = 2950a
  • Abbreviations: APN, advanced practice nurse; EHR, electronic health record system; FTE, full-time equivalent; HF, heart failure.

  • a

    P < 0.0001 for all comparisons. P values for continuous variables were based on analysis of variance; P values for categoric values were based on χ2 tests.

Region, no. (%)   
 South1595 (37.8)244 (33.7)1228 (41.6)
 West756 (17.9)13 (1.8)255 (8.6)
 Central836 (19.8)149 (20.6)477 (16.2)
 Northeast1033 (24.5)317 (43.8)990 (33.6)
Practice setting, no. (%)   
 Nonuniversity, nonteaching2816 (66.7)410 (56.7)2043 (69.3)
 Nonuniversity, teaching1003 (23.8)264 (36.5)589 (20.0)
 University, teaching401 (9.5)49 (6.8)318 (10.8)
Multispecialty practice, no. (%)1015 (24.1)74 (10.2)827 (28.0)
>1 APN (FTE) on staff, no. (%)2118 (50.9)292 (40.4)903 (31.9)
Electrophysiologists in practice, median (25th–75th percentile)2 (1–3)1 (0–3)1 (0–2)
Interventionalists in practice, median (25th–75th percentile)5 (3–6)5 (4–7)4 (3–6)
Heart failure clinic in practice, median (25th–75th percentile)1 (1–2)2 (2–2)2 (1–2)
Cardiologists in practice, median (25th–75th percentile)12 (8–20)11 (7–14)10 (6–17)
No. of HF patients managed annually in practice, median (25th–75th percentile)2500 (900–5850)2125 (800–2500)1500 (600–5000)

Table 2 presents demographic and clinical characteristics of patients in the 3 practice groups. The median age of patients in each group was similar, and a majority of patients in all 3 groups were male. NYHA class III or IV was 21.0% in EHR-always practices, 13.0% in EHR-switched practices, and 20.9% in paper-always practices (P < 0.0001). The paper-always group had the highest proportion of patients with diabetes. History of MI did not differ between groups.

Table 2. Baseline Patient Characteristics by Type of Medical Record System
CharacteristicEHR-Always, n = 4220EHR-Switched, n = 723Paper-Always, n = 2950P
  1. Abbreviations: EHR, electronic health record system; NYHA, New York Heart Association.

Age, median (25th–75th percentile), y70 (60–77)70 (60–77)69 (59–77)0.3193
Male, no. (%)3024 (71.7)495 (68.5)2111 (71.6)0.1977
Race, no. (%)    
 White1696 (40.9)379 (52.8)1281 (44.4) 
 Other412 (9.9)82 (11.4)376 (13.0) 
 Not documented/missing2041 (49.2)257 (35.8)1231 (42.6)0.0174
Ethnicity, no. (%)    
 Hispanic/Latino80 (2.0)9 (1.2)61 (2.1) 
 Not Hispanic/Latino487 (12.0)66 (9.2)306 (10.8) 
 Not documented3475 (86.0)646 (89.6)2471 (87.1)0.4412
Insurance    
 Medicare2526 (60.0)380 (52.9)1718 (58.5) 
 Other1493 (35.5)239 (33.2)1028 (35.0) 
 Not documented190 (4.5)100 (13.9)192 (6.5)0.7803
Heart failure etiology, ischemic, no. (%)2795 (66.5)456 (63.3)1894 (64.7)0.1210
History of atrial fibrillation/flutter, no. (%)1301 (30.8)220 (30.4)842 (28.5)0.1096
History of coronary artery bypass graft, no. (%)1289 (30.5)235 (32.5)933 (31.6)0.4383
History of percutaneous coronary intervention, no. (%)1108 (26.3)181 (25.0)775 (26.3)0.7739
History of myocardial infarction, no. (%)1713 (40.6)281 (38.9)1267 (42.9)0.0511
History of peripheral vascular disease, no. (%)388 (9.2)74 (10.2)303 (10.3)0.2769
History of diabetes, no. (%)1301 (30.8)231 (32.0)994 (33.7)0.0378
History of hypertension, no. (%)2607 (61.8)429 (59.3)1823 (61.8)0.4349
History of chronic obstructive pulmonary disease, no. (%)595 (14.1)107 (14.8)475 (16.1)0.0641
History of depression, no. (%)355 (8.4)78 (10.8)229 (7.8)0.0313
NYHA class documented4123 (97.7)711 (98.5)2899 (98.3)0.1059
NYHA class    
 I or II3236 (76.7)617 (85.5)2284 (77.5) 
 III or IV887 (21.0)94 (13.0)615 (20.9) 
 Not documented97 (2.3)11 (1.5)49 (1.7)<0.0001
Left ventricular ejection fraction, median (25th–75th percentile), %25 (20–30)25 (20–30)25 (20–30)0.7742
Systolic blood pressure, median (25th–75th percentile), mm Hg120 (108–130)120 (109–132)120 (110–132)0.0001
Diastolic blood pressure, median (25th–75th percentile), mm Hg70 (62–78)70 (60–78)70 (62–80)0.0002
Resting heart rate, median (25th–75th percentile), beats/min70 (64–78)72 (64–78)70 (64–78)0.3670
Rales on most recent exam, no. (%)130 (3.1)11 (1.5)76 (2.6)0.1386
Edema on most recent exam, no. (%)744 (17.8)119 (16.6)433 (14.9)0.0012
Sodium, median (25th–75th percentile), mEq/L140 (138–142)140 (137–142)140 (138–142)0.9413
Blood urea nitrogen, median (25th–75th percentile), mg/dL21 (16–28)21 (16–28)21 (16–28)0.3679
Creatinine, median (25th–75th percentile), mg/dL1.2 (1.0–1.5)1.2 (1.0–1.5)1.2 (1.0–1.4)0.2018
Potassium, median (25th–75th percentile), mEq/L4.4 (4.1–4.7)4.4 (4.1–4.6)4.4 (4.1–4.7)0.1318
B-natriuretic peptide, median (25th–75th percentile), pg/mL299.5 (130–672)301 (135–717)324.5 (137–687)0.9824
QRS duration, median (25th–75th percentile), ms124 (100–156)130 (102–158)125 (100–156)0.5822
QRS missing, no. (%)1323 (31.4)340 (47.0)826 (28.0)<0.0001

A significant improvement in use of 6 of the 7 quality measures was demonstrated at 24 months after implementation of the intervention overall (P < 0.001). In this study, EHR-always practices demonstrated significant improvement in 5 of the 7 measures (β-blocker, aldosterone antagonist, CRT, ICD, and HF education), no change in 1 measure (ACEI/ARB), and worsening in 1 measure (anticoagulation for atrial fibrillation) (Figure 1). EHR-switched and paper-always practices each had significant improvement in 6 of 7 measures and no change in the anticoagulation measure (Figure 1). Table 3 shows the results of univariate and multivariate analyses to determine the association between type of medical record system and magnitude of improvement in use of the 7 quality measures from baseline to 24 months. There were no differences in the magnitude of improvement between EHR-always, EHR-switched, and paper-always practices for any of the individual measures, with only 1 exception. A statistically significantly greater improvement in use of β-blockers in eligible patients was detected in the EHR-always group as compared with the paper-always group (adjusted odds ratio [95% confidence interval]: 1.43 [1.05–1.93]; P = 0.0225). All other differences between groups were nonsignificant. Practice-level analyses also demonstrated similar magnitudes of improvement in quality measures irrespective of practice use of EHR for most measures (Supplemental Table 1). EHR-always practices showed no greater absolute improvement compared with paper-always practices for any individual measure, except for the ICD measure. There was also greater absolute improvement in use of aldosterone antagonists (P = 0.0226) in EHR-switched practices as compared with paper-always practices (Supplemental Table 1).

thumbnail image

Figure 1. Rates of adherence to 7 quality measures at baseline and 24 months for (A) electronic health record systems (EHR)-always, (B) EHR-switched, and (C) paper-always practices. All patients eligible for an individual quality measure at baseline or at 24 months were included. Abbreviations: ACEI/ARB, angiotensin-converting enzyme inhibitor/ angiotensin receptor blocker; CRT, cardiac resynchronization therapy; EHR, electronic health record; HF, heart failure; ICD, implantable cardioverter-defibrillator.

Download figure to PowerPoint

Table 3. Magnitude of Improvement in Use of Evidence-Based Heart Failure Therapies From Baseline to 24 Months by Type of Medical Record Systema
  • Abbreviations: ACEI/ARB, angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; CI, confidence interval; EHR, electronic health record system; HF, heart failure; OR, odds ratio.

  • a

    Improvement for each quality measure was defined as new treatment at 24 months of patients who were eligible for the quality measure at both time points but who were not treated at baseline. Patients who were ineligible at either time point were excluded.

    Univariate AnalysisMultivariate Analysis
 Improvement in Use of Therapy by EHR Group (Absolute Rate, Unadjusted), %EHR-Always vs Paper-AlwaysEHR-Switched vs Paper-AlwaysEHR-Always vs Paper-AlwaysEHR-Switched vs Paper-Always
Quality MeasureEHR-Always %EHR-Switched %Paper-Always %Unadjusted OR (95% CI)PUnadjusted OR (95% CI)PAdjusted OR (95% CI)PAdjusted OR (95% CI)P
ACEI/ARB7.39.38.60.83 (0.64–1.07)0.15381.03 (0.68–1.55)0.89070.83 (0.63–1.09)0.18351.02 (0.66–1.57)0.9415
β-Blocker6.97.75.31.24 (0.90–1.69)0.18371.30 (0.79–2.15)0.30631.43 (1.05–1.93)0.02251.65 (1.00–2.74)0.0506
Aldosterone antagonist17.427.820.70.83 (0.55–1.25)0.36731.45 (0.65–3.27)0.36420.86 (0.49–1.50)0.58930.97 (0.36–2.61)0.9490
Anticoagulation for atrial fibrillation6.44.98.60.71 (0.50–1.01)0.05510.53 (0.31–0.92)0.02470.65 (0.40–1.05)0.07950.73 (0.35–1.54)0.4105
Implantable cardioverter-defibrillator19.116.318.01.09 (0.87–1.36)0.44670.82 (0.56–1.22)0.33411.06 (0.78–1.44)0.69360.67 (0.42–1.07)0.0901
Cardiac resynchronization therapy33.641.731.11.15 (0.69–1.92)0.58891.60 (0.63–4.09)0.32751.33 (0.73–2.43)0.35270.81 (0.26–2.53)0.7199
HF education24.731.426.60.83 (0.60–1.15)0.26061.31 (0.86–2.00)0.20740.95 (0.67–1.35)0.78881.40 (0.88–2.22)0.1532

Table 4 shows the associations between type of medical record system and adherence to the quality measures at 24 months for the 3 groups. After implementation of the performance-improvement initiative, there was no significantly greater use of guideline-recommended HF care among EHR-always or EHR-switched practices over the course of the study compared with paper-always practices for any of the quality measures. In practice-level analyses, adherence to evidence-based therapy at 24 months was similar between practices using EHR-always, EHR-switched, and paper-always record systems for all but 1 measure (Supplemental Table 2). Use of an ICD in eligible patients at 24 months was higher among EHR-always practices compared with paper-always practices.

Table 4. Association Between Type of Medical Record System and Adherence to Evidence-Based Therapy at 24 Monthsa
  • Abbreviations: ACEI/ARB, angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; CI, confidence interval; EHR, electronic health record system; HF, heart failure; OR, odds ratio.

  • a

    Univariate and multivariate generalized estimating equation models were used to analyze associations for the quality measures.

    Univariate AnalysisMultivariate Analysis
 Adherence to Therapy at 24 Months (Unadjusted), %EHR-Always vs Paper-AlwaysEHR-Switched vs Paper-AlwaysEHR-Always vs Paper-AlwaysEHR-Switched vs Paper-Always
Quality MeasureEHR-Always %EHR-Switched %Paper-Always %Unadjusted OR (95% CI)PUnadjusted OR (95% CI)PAdjusted OR (95% CI)PAdjusted OR (95% CI)P
ACEI/ARB85.386.285.51.05 (0.82–1.33)0.70261.07 (0.73–1.57)0.72251.25 (0.93–1.69)0.13901.07 (0.65–1.78)0.7816
β-Blocker92.991.392.41.18 (0.87–1.62)0.29230.94 (0.62–1.44)0.78540.98 (0.64–1.49)0.93151.07 (0.56–2.05)0.8356
Aldosterone antagonist61.759.855.81.41 (1.04–1.91)0.02581.51 (0.89–2.57)0.12720.94 (0.51–1.75)0.84891.05 (0.44–2.50)0.9196
Anticoagulation for atrial fibrillation70.172.665.31.14 (0.89–1.47)0.30591.25 (0.76–2.05)0.37291.17 (0.84–1.64)0.35971.11 (0.74–1.68)0.6021
Implantable cardioverter-defibrillator78.079.879.21.07 (0.80–1.42)0.66321.21 (0.68–2.15)0.52051.10 (0.79–1.53)0.58801.51 (0.86–2.67)0.1526
Cardiac resynchronization therapy64.975.869.00.81 (0.57–1.15)0.23261.48 (0.71–3.08)0.29881.07 (0.68–1.67)0.78152.25 (0.60–8.40)0.2278
HF education75.058.565.91.62 (1.09–2.42)0.01820.85 (0.50–1.44)0.53361.51 (0.98–2.32)0.06250.85 (0.51–1.44)0.5528

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

In this study of outpatient cardiology practices from across the United States, use of EHR was not associated with additional improvement in use of guideline-recommended therapies for HF after implementation of a performance-improvement system compared with use of paper-only records. We found, with a single exception, no association between EHR use throughout the study or switching to EHR systems and greater improvements in HF care quality for 7 measures compared to practices that remained paper-based. The quality of care at 24 months was also similar among the 3 practice groups. Although some studies suggest that EHR can facilitate better quality of care, the lack of associations between EHR and improvements in quality in our study and other recent studies17 fails to support such conclusions, suggests that EHR systems as currently deployed do not facilitate improvement in HF care quality, and has important implications for the healthcare system.

Heart failure is a chronic disease that results in more than 12 million outpatient office visits annually as well as substantial morbidity and mortality. Although there are a number of evidence-based, mortality-reducing therapies for HF and national guidelines that provide recommendations for use of these therapies, many eligible HF patients do not receive these recommended therapies, and substantial variations in HF care by outpatient practices have been reported. Health information technologies (HIT) that support disease management by clinicians could significantly improve quality of care and clinical outcomes of patients with HF. The American Reinvestment and Recovery Act set aside $19.2 billion to promote HIT use in the United States, based on the presumption that deployment of EHR will provide for meaningful improvements in the quality of care. The majority of HITECH funding goes to raising reimbursement rates for Medicare and Medicaid services delivered with meaningful use of EHR. A survey of EHR use conducted in 2009 by the National Center for Health Statistics revealed that 48.3% of physicians reported use of full or partial EHR systems.18 Prior studies have reported low rates of EHR adoption among outpatient practices, including single-specialty and multispecialty practices. At the initiation of this study, 50.3% of cardiology practices were utilizing EHR systems. During the course of the 24-month study, an additional 9.7% of practices switched from paper-based systems to EHR systems.

In the present study, we observed no difference in the magnitude of improvement in use of HF quality measures over time between practices with and without EHR, even in the setting of an overall successful performance-improvement initiative. As such, our findings raise significant doubts about the ability of broad EHR system adoption as promoted by HITECH to independently improve the quality of outpatient care. In the absence of evidence supporting greater improvement in quality of care with EHR, it is possible that this planned investment would be better devoted to interventions, such as performance-improvement programs like IMPROVE HF, proven to improve quality of care and clinical outcomes.

We previously reported little to no difference in HF care quality among IMPROVE HF practices at baseline with regard to differential use of EHR, prior to implementation of the performance-improvement initiative.13 Other studies have also failed to find a difference in care quality with EHR system use.19,20 Linder and colleagues6 found no differences in overall EHR use and quality of care in a national ambulatory data set. Using data from the 2005–2007 National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey, Romano and Stafford17 analyzed 255 402 ambulatory patient visits to assess the relationship of EHR and clinical decision support to the provision of guideline-concordant care and found no differences in quality of care for 19 of 20 measures. Further, when the investigators compared a subset of advanced-function EHR systems with more basic, limited-function EHRs, they still found no difference in quality. Our analysis extends the findings of these prior studies by demonstrating that even in the setting of a focused and overall successful performance-improvement initiative, there were no greater improvements in care quality among practices using EHR for the majority of quality measures.

Current EHR systems vary enormously in their functionality. Current systems may not be sufficient to foster greater improvement in care compared with paper-based medical records. The level of decision support available in the EHR systems reported by IMPROVE HF practices varied widely, with some practices reporting systems with the potential for a completely customized platform and others using systems limited to document scanning only.13 Heterogeneity of EHR usage among participating practices prevents correlations between specific features of various EHR systems and conformity to recommended measures. It is possible that larger, more consistent improvements in the use of all guideline-recommended therapies might have been observed if greater uniformity of EHR decision support was available to practices participating in IMPROVE HF. Substantial revamping of EHR systems with better integration of clinical decision support, real-time performance feedback, and other enhancements may be required. These findings may suggest a need for even greater attention to coordinated implementation of clinical decision support to realize the full potential of EHR systems coupled to performance-improvement initiatives to improve healthcare quality beyond that which can be achieved with performance-improvement systems alone.

There remains an important need to better understand the effects of EHR systems on clinical care to ensure that healthcare organizations, clinicians, patients, and policymakers can make informed decisions about where best to invest healthcare resources to improve the quality of patient care. Future research should investigate why the use of, or conversion to, EHR systems has not translated into meaningful quality improvement even when coupled with well-validated and successful performance-improvement initiatives as well as investigate ways to ensure the meaningful improvement in quality results with use of EHR systems. This information will be vital to better focus HIT implementation efforts such that a clinically relevant impact on care quality and outcomes can be achieved. It is also essential to ensure the efforts to implement EHR systems do not divert attention from and investment in performance-improvement programs that have been proven to be clinically effective.

Limitations

Data were collected by medical chart review. Thus, the quality and validity of these data depend on the accuracy and completeness of the medical records and abstraction process. It is possible that changes in treatment rates found in this study may be attributable in part to variations or inaccuracies in the medical record or data abstraction process. As expected in the clinical practice setting, follow-up was not available for all patients. This incomplete follow-up may have introduced a selection bias that may have influenced some or all of the findings, although the baseline characteristics and treatment rates of patients with and without complete follow-up were previously shown to be similar.15 Although use of aldosterone antagonists, CRT, and ICD therapy have been shown to reduce mortality and are Class I recommendations for eligible patients in the ACC/AHA guidelines, the outpatient performance measure sets do not include measures for these therapies. It is possible that the improvements in quality measures may have been influenced by secular trends and factors other than participation in this study. It is also possible that the improvements in care resulting from the IMPROVE HF performance-improvement initiative were so large that they limited the ability to discern small, but still clinically meaningful, differences in care quality resulting from use or conversion to EHR systems. Residual measured and unmeasured confounding variables may have influenced some or all of the findings. The practices participating in IMPROVE HF were self-selected; thus, these findings may not apply to practices that differ in patient case mix, baseline care patterns, motivation, and resources. Use of EHR in this study may not accurately represent their use in other outpatient practices, and findings from this study may not extrapolate to cardiology practices not included in the IMPROVE HF cohort and disease states other than HF.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

Implementation of the defined and scalable IMPROVE HF practice-specific performance-improvement intervention enhanced use of guideline-recommended HF therapies among cardiology and multispecialty practices irrespective of EHR use. However, despite the promise of better quality, cardiology practices utilizing or newly converting to EHR systems did not achieve greater improvements in the quality of HF care compared with practices using paper-based systems. These findings raise doubts about whether implementation of costly EHR technologies will translate into better outpatient quality of care. Efforts and resources may be better devoted to fostering greater implementation of systems proven to improve the quality of care and outcomes.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information

CommGeniX, LLC (Tampa, FL), provided technical support with funding from Medtronic, Inc.

Individual author disclosures are as follows.

Dr. Walsh: consultant for Medtronic (modest), United HealthCare (modest), BioControl (modest), and EMERGE (modest).

Dr. Albert: No relationships to disclose.

Dr. Curtis: research grants from Medtronic (modest) and St. Jude Medical (modest); speakers bureau for Medtronic (modest) and Sanofi-Aventis (significant); consultant/advisory board for Sanofi-Aventis (modest), Medtronic (modest), St Jude Medical (modest), and Biosense Webster (modest);fellowship support from Medtronic (significant).

Dr. Gheorghiade: consultant for Abbott Laboratories (modest),Astellas (modest), AstraZeneca (modest), Bayer Schering Pharma AG (significant), CorThera, Inc (modest),Cytokinetics, Inc (modest), DebioPharm SA (significant), Errekappa Terapeutici (Milan, Italy) (modest), GlaxoSmithKline (modest), Johnson & Johnson (modest), Medtronic (significant), Merck (modest), Novartis Pharma AG(significant, Otsuka Pharmaceuticals (significant, Pericor Therapeutics (significant), Protein Design Laboratories (modest), Sanofi-Aventis (modest), Sigma Tau (significant), and Solvay Pharmaceuticals (significant).

Dr. Heywood: research grants from Biosite (modest), Medtronic (significant), and St Jude (significant); speakers bureau/honoraria from GlaxoSmithKline (modest), Medtronic (significant), AstraZeneca (modest), Novartis (modest), Actelion (significant), St Jude (modest), Otsuka (modest), and Boston Scientific (modest); consultant/advisory board for Emerge (modest), Medtronic (significant), and Actelion (modest. Ms. Liu: employee of Medtronic. Dr. Mehra: consultant for Medtronic (modest), Johnson & Johnson (modest), and St Jude Medical (modest); grants/research support from the Maryland Tobacco Fund (significant) and National Institutes of Health/NHLBI (significant). Dr. O'Connor: consultant for Forest (significant), Medtronic (significant), Amgen (significant), Medpace (significant), Impulse Dynamics (significant), Actelion (significant), Cytokinetics (modest), Roche (modest), and Trevena (modest).

Dr. Reynolds: research grants from Medtronic and Biotronik; speakers bureau for Medtronic, Sorin, and St Jude Medical; consultant for Medtronic (all significant). Dr. Yancy: No relationships to disclose.

Dr. Fonarow: research grants from the National Institutes of Health (significant); consultant for Medtronic (significant), Novartis (significant), and St Jude's (modest); honoraria from Medtronic (significant), GlaxoSmithKline (significant), Pfizer (significant), and St Jude's (modest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information
  • 1.
    Balfour DC III, Evans S, Januska J, et al. Health information technology—results from a roundtable discussion. J Manag Care Pharm. 2009;15(1 suppl A):S10S17.
  • 2.
    Jha AK. Meaningful use of electronic health records: the road ahead. JAMA. 2010;304:17091710.
  • 3.
    Dick RS, Steen EB, Detmer DE, eds; Committee on Improving the Patient Record, Institute of Medicine. The Computer-Based PatientRecord: An Essential Technology for Health Care, Revised Edition. Washington, DC: National Academy Press; 1997.
  • 4.
    Committee on Quality of Health Care in America, Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press; 2001.
  • 5.
    Poon EG, Wright A, Simon SR, et al. Relationship between use of electronic health record features and health care quality: results of a statewide survey. Med Care. 2010;48:203209.
  • 6.
    Linder JA, Ma J, Bates DW, et al. Electronic health record use and the quality of ambulatory care in the United States. Arch Intern Med. 2007;167:14001405.
  • 7.
    Welch WP, Bazarko D, Ritten K, et al. Electronic health records in four community physician practices: impact on quality and cost of care. J Am Med Inform Assoc. 2007;14:320328.
  • 8.
    Garrido T, Jamieson L, Zhou Y, et al. Effect of electronic health records in ambulatory care: retrospective, serial, cross sectional study. BMJ. 2005;330:581.
  • 9.
    Weber V, Bloom F, Pierdon S, et al. Employing the electronic health record to improve diabetes care: a multifaceted intervention in an integrated delivery system. J Gen Intern Med. 2007;23:379382.
  • 10.
    Riggio JM, Sorokin R, Moxey ED, et al. Effectiveness of a clinical-decision-support-system in improving compliance with cardiac-care quality measures and supporting residency training. Acad Med. 2009;84:17191726.
  • 11.
    Chaudhry B, Wang J, Wu S, et al. Systematic review: impact of health information technology on quality, efficiency, and costs of medical care. Ann Intern Med. 2006;144:742752.
  • 12.
    Garg AX, Adhikari NK, McDonald H, et al. Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: a systematic review. JAMA. 2005;293:12231238.
  • 13.
    Walsh MN, Yancy CW, Albert NM, et al. Electronic health records and quality of care for heart failure. Am Heart J. 2010;159: 635642.
  • 14.
    Fonarow GC, Yancy CW, Albert NM, et al. Improving the use of evidence-based heart failure therapies in the outpatient setting: the IMPROVE HF performance improvement registry. Am Heart J. 2007;154:1238.
  • 15.
    Fonarow GC, Albert NM, Curtis AB, et al. Improving evidence-based care for heart failure in outpatient cardiology practices: primary results of the Registry to Improve the Use of Evidence-Based Heart Failure Therapies in the Outpatient Setting (IMPROVE HF). Circulation. 2010;122:585596.
  • 16.
    Bonow RO, Bennett S, Casey DE Jr, et al. ACC/AHA clinical performance measures for adults with chronic heart failure: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures (Writing Committee to Develop Heart Failure Clinical Performance Measures): endorsed by the Heart Failure Society of America. J Am Coll Cardiol. 2005;46:11441178.
  • 17.
    Romano MJ, Stafford RS. Electronic health records and clinical decision support systems: impact on national ambulatory care quality. Arch Intern Med. 2011;171:897903.
  • 18.
    Hsiao CJ, Hing E, Socey TC, et al. Electronic medical record/ electronic health record systems of office-based physicians: United States, 2009 and preliminary 2010 state estimates. Atlanta, GA: National Center for Health Statistics; 2010. http://www.cdc.gov/nchs/data/hestat/emr_ehr_09/emr_ehr_09.htm. Accessed March 2011.
  • 19.
    Keyhani S, Hebert PL, Ross JS, et al. Electronic health record components and the quality of care. Med Care. 2008;46:12671272.
  • 20.
    Zhou L, Soran CS, Jenter CA, et al. The relationship between electronic health record use and quality of care over time. J Am Med Inform Assoc. 2009;16:457464.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
clc_21971_sm_supportinginfo.doc67KSupplemental Table 1. Absolute Improvement in Quality Measures From Baseline to 24 Months by Type of Medical Record System: Practice-Level Analyses

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.