|Home | About | Journals | Submit | Contact Us | Français|
Registry studies have suggested improvements in door-to-balloon times, but a national assessment of the trends in door-to-balloon times is lacking. Moreover, we do not know if improvements in door-to-balloon times were shared equally among patient and hospital groups.
This analysis includes all patients reported by hospitals to the Centers for Medicare & Medicaid Services for inclusion in the time to percutaneous coronary intervention acute myocardial infarction-8 (AMI-8) inpatient measure from January 1, 2005 through September 30, 2010. For each calendar year, we summarized the characteristics of patients reported for the measure including the number and percentage in each group as well as the median time to primary percutaneous coronary intervention, and the percentage with time to primary percutaneous coronary intervention within 75 minutes and within 90 minutes. Door-to-balloon time declined from a median of 96 minutes in the year ending December 31, 2005 to a median of 64 minutes in the three quarters ending September 30, 2010. There were corresponding increases in the percentage of patients who had times <90 minutes (44.2% to 91.4%) and <75 minutes (27.3% to 70.4%). The declines in median times were greatest among groups that had the highest median times during the first period: Patients >75 years of age (median decline 38 minutes); women (35 minutes); and African Americans (42 minutes).
National progress has been achieved in the treatment of patients with ST-segment elevation myocardial infarction who undergo primary percutaneous coronary intervention.
The effectiveness of primary percutaneous coronary intervention (PCI) is highly dependent on its timeliness.1-4 Clinical practice guidelines and practice guidelines recommend that the time from hospital arrival to mechanical reperfusion, the door-to-balloon (D2B) time, should be as short as possible and not exceed 90 minutes.5-6 In 2002, only a third of patients received primary PCI within 90 minutes and a third underwent the procedure more than 2 hours after arriving at the hospital.7 These lackluster times led to a national response; in September 2005, the Centers for Medicare & Medicaid Services (CMS) began to publicly report the percent of patients treated within recommended times. At the same time, federally-funded research identified strategies and organizational factors that were strongly associated with shorter D2B times.8-11 In November 2006, the American College of Cardiology (ACC), with national partners, launched the D2B Alliance, a national campaign to improve D2B times by advocating the adoption of key strategies that had been shown to reduce delays based on a study funded by the National Heart, Lung, and Blood Institute.12 In May 2007, the American Heart Association (AHA) launched Mission: Lifeline, another national initiative to improve systems of care for patients with ST-segment elevation myocardial infarction (STEMI).13
Although studies have reported improvements in D2B times during this period of national focus on improving the timeliness of primary PCI,14-15 these reports, which assessed performance through the first quarter of 2008, were derived from hospitals participating in registries, which represent a selected sample of the nation’s hospitals. There has been no national assessment of the trends in D2B times, nor do we know if improvements in D2B times were shared equally among patient and hospital groups.
Accordingly, we evaluated data submitted to CMS as part of its initiative to report indicators of quality for patients with an acute myocardial infarction (AMI). The latest published report from the CMS data included data up until the second quarter of 2004.16 We sought to determine how performance on this measure had changed from 2005 through 2010, with a focus on D2B times as well as the percent of patients who were treated in <90 minutes and in <75 minutes. We also evaluated the trends among hospitals defined by their bed size, geographic region, ownership, urban or rural location, and the volume of patients with PCI that they reported for this measure. This study represents the most contemporary, comprehensive, nationally representative investigation of the changes in D2B times in the United States.
This analysis includes all patients reported by hospitals to CMS for inclusion in the time to PCI (AMI-8) inpatient measure from January 1, 2005 through September 30, 2010. Hospitals with at least 5 AMI inpatient admissions during a quarter must report to CMS or face a financial penalty. Hospitals may exclude patients through sampling, depending on the number of patients eligible, and the minimum number required to be reported has changed over time. Sampling is mandated by CMS to be either simple random or sequential random, and hospitals that sample for at least one quarter must report that they sampled for that year; this is enforced through an additional mandate to use a measurement system certified by The Joint Commission.17 The percentage of hospitals that sampled was generally stable over the study period, with the largest number (percentage) being 252/4085 (6.2%) in 2005 and the smallest number being 173/3780 (4.6%) in 2010. The inclusion and exclusion criteria for this measure, as well as the sampling and reporting criteria, are publicly available.17 During the period of the study, there were some changes in the measure regarding exclusion criteria; the major change occurred in 2006 when patients were allowed to be excluded if there was a non-clinical reason for the delay (AMI-8 exclusions: online supplement). To increase the likelihood that we were assessing patients receiving primary PCI, we excluded patients with times >6 hours.
For each patient included in the measure, there is information about age, sex, and racial/ethnic group (white, black, other, unknown).
We classified each hospital that reported at least 25 patients during any year for the measure according to size (number of beds), Census region, ownership (government, forprofit, and nonprofit), location (rural or urban), and number of patients treated with PCI, as submitted by the hospital. For 2010, we repeated the analysis with hospitals that reported at least 20 patients because we only had three quarters of data. Information about hospitals was taken from the Program Resource System, a national provider database maintained by CMS and the Quality Improvement Organizations.
For each year, we summarized the characteristics of patients, including the number and percentage in each group as well as the median time to primary PCI, and the percentage with time to primary PCI within 75 minutes and within 90 minutes for each group.
For each year, we summarized hospital characteristics, number of patients reported, and percentage of hospitals in each group. For each year, we also calculated the median and range of hospital median time to PCI, average of hospital percentage of patients treated within 90 minutes and within 75 minutes, and the interquartile range (IQR) for each of these. We graphed the summary measure of percent <90 minutes and <75 minutes over the 6-year period for all patients included in the study.
We conducted the analyses with SAS version 9.1.3 [2004. SAS Institute Inc., Cary, NC, USA] and Stata version 11.1 [2010. Stata Corp, College Station, TX, USA]. The Human Investigation Committee at the Yale University School of Medicine approved an exemption for the authors to use CMS claims and enrollment data for research analyses and publication; informed consent was not required.
During the 6-year period, the number of patients was fairly constant, ranging between 48,977 and 53,682 (Table 1), with 42,150 reported during the 3 quarters of 2010. The number of hospitals that reported at least 25 patients increased slightly from 896 to 973, with 764 reporting 25 patients for the first 3 quarters of 2010 (Table 2).
Median D2B times declined 32 minutes over the 6-year period from a median of 96 minutes in the year ending December 31, 2005 to a median of 64 minutes in the year ending September 30, 2010 (Table 1). The declines in median times were greatest among groups that had the highest median times during the first period: Patients >75 years of age (median decline 38 minutes); women (35 minutes); and African Americans (42 minutes). There were corresponding increases in the percentage of patients who had D2B times of <90 minutes (44.2% to 91.4%) and <75 minutes (27.3% to 70.4%) (Figures 1 and and22).
Hospital median D2B times declined over the 6-year period from a median of 97 minutes in 2005 to a median of 64 minutes in 2010 (Table 2). The declines were greatest among groups of hospitals that had the longest times in the first year: Hospitals with 500 or more beds (median hospital median declined 34 minutes); for-profit hospitals (declined 38 minutes); and hospitals in the East South Central and Mid Atlantic Census regions (40- and 35-minute declines, respectively). There was a corresponding increase in the hospital average percent of patient D2B times <90 minutes (from 44% to 92%) and <75 minutes (from 27% to 71%). The results from 2010 were similar if we restricted the required number of cases to 20.
The study demonstrates the national progress in the treatment of patients with STEMI who undergo primary PCI. Median patient D2B times decreased substantially from 96 minutes to 64 minutes, a drop of 32 minutes, over the 6 years ending in mid-2010, representing more than a 30% relative decline. This improvement, experienced across the country and across different types of hospitals, represents a remarkable elevation in practice that was achieved over a relatively short period of time and in the absence of financial incentive. The accomplishment is truly a tribute to interventional cardiologists, emergency medicine physicians, nurses, technologists, and other team members nationwide who were dedicated to improving D2B times.
The perspective on D2B times changed dramatically over this period. The 2004 STEMI guidelines recommended that patients be treated with primary PCI within 90±30 minutes.18 The caveat of the additional 30 minutes was included as a compromise in response to controversy about whether it was possible for hospitals to routinely treat patients with STEMI within 90 minutes. The publication of the 2004 guidelines was followed by a shift toward the 90-minute standard. Nallamothu and colleagues published a study based on trials of the relationship between D2B time and the advantage of primary PCI over fibrinolytic therapy.1 They found that if primary PCI was delayed more than an hour beyond the time that fibrinolytic therapy could be provided, the advantage was lost. Thus, if fibrinolytic therapy was recommended to be given within 30 minutes, then the provision of primary PCI within 90 minutes was supported by evidence from the trials. In the guideline update published in 2007, the additional 30 minutes was removed, altering the recommendation for D2B time to <90 minutes.6 The CMS measure, which initially reported the percent of patients treated within 120 minutes, was subsequently reduced to 90 minutes in 2006 to align the performance measure with the new guideline recommendation.17
The improvement in D2B times that we observed cannot be definitively attributed to any single action, but many activities likely contributed. During this period, multiple national efforts focused attention on timeliness of D2B and supported quality improvement. Published articles that revealed gaps in care and indicated strategies that were associated with faster times contributed to clinical changes in performance. A study sponsored by the National Institutes of Health used a mixed methods approach to examine exceptional performers and then test hypotheses that derived from their experience.7-10,19 CMS developed contracts with Quality Improvement Organizations that contributed to the increasing focus on improving various aspects of AMI care, including D2B times. Hospital groups and consortia focused on improving D2B times. The performance of the nation’s hospitals in treating patients with STEMI was further highlighted by the release of the publicly reported D2B time measures by CMS, and the D2B Alliance and Mission: Lifeline, national campaigns by the ACC and the AHA, enlisted clinicians and hospitals in a broadbased effort to reduce delays.12-13
Improvement in D2B times demonstrates the way that emerging science about how to improve care can be rapidly integrated into practice. The ACC campaign was launched simultaneously with the publication of a paper in the New England Journal of Medicine that described strategies associated with faster times.9 The ACC campaign promoted the adoption of such strategies, which were shown to be underutilized nationally. Recent reports demonstrated the marked integration of these strategies into practice that occurred during the period of the campaign.14 Moreover, a recent qualitative study showed that the credibility of the campaign was related to the strength of the science and the clarity of the recommendations.20
Despite the recent gains, additional opportunities for improvement in D2B times remain. The most outstanding institutions are now regularly achieving exceptional times of approximately 60 minutes through strategies including coordination with Emergency Medical Services and the collection and dissemination of a pre-hospital electrocardiogram.21-23 This level of performance may become the new standard.
Another opportunity for improvement is related to the care of transfer patients. Prior studies have shown that many patients who are transferred from a hospital without PCI capability to a PCI-capable institution experience long delays in treatment.24-25 To address this concern, CMS is collecting a measure, with potential to be publicly reported, which assesses the time required to transfer such patients.17 More importantly, current research shows that these times can be reduced through greater coordination between hospitals.26-27
A limitation of this assessment is the evolution of the measure over the study period, with modifications primarily related to the exclusion criteria. The most notable change occurred in 2006 and allowed hospitals to exclude patients based on the judgment that a D2B time >90 minutes was the result of a delay incurred by patient preference or clinical condition. Subsequent changes included minor alterations in the codes or slight expansions in the exclusion criteria. However, the dramatic decline in D2B times that was observed over the study period is unlikely to have resulted from changes in the measure. Moreover, the greatest decline occurred between 2006 and 2007, a period corresponding to the initiation of national campaigns to improve D2B times. Finally, our results are aligned with those of registries that documented trends in D2B times and applied consistent criteria over time.
In conclusion, we document remarkable improvement in D2B times from 2005 through 2010. The improvement demonstrates the results that can be produced by collaboration between health care professionals, hospitals, federal research agencies and national organizations interested in patient care toward the achievement of a shared goal. The focus on improving the way in which care is delivered - improving the systems - has yielded more timely care for patients and serves as a template for similar contemporary and future efforts in areas such as readmission.
The authors thank Dr. Allen Ma from the Oklahoma Foundation for Medical Quality for his contributions to this manuscript.
Funding Sources: This work was conducted under a federal government contract with the Centers for Medicare & Medicaid Services (CMS) - Contract HHSM-500-2008-00025I, Task Order T0001. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. The authors assume full responsibility for the accuracy and completeness of the ideas presented. CMS staff fully participated in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and critical review of the manuscript. CMS reviewed and approved the use of its data for this work and approved submission of the manuscript.
Drs. Krumholz and Curtis are supported by grant U01 HL105270 (Center for Cardiovascular Outcomes Research at Yale University) from the National Heart, Lung, and Blood Institute.
DisclosuresDr. Krumholz reports that he chairs a cardiac scientific advisory board for UnitedHealth. The other authors have no potential financial conflicts to disclose.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.