The emergence of information technology in healthcare holds the promise to transform the industry through the creation of highly reliable information exchange. These same technologies have a central role in the patient safety movement. Organizations that wish to deliver safe and high-quality healthcare will only be successful if they plan, develop, and use health information systems with the principles of high-performing organizations in mind. We discuss the current state of health information technology in the patient safety movement, how this technology can contribute to high organizational performance, and some caveats.

Healthcare organizations can advance patient safety by using the principles of high-performing organizations when they design and implement health information technology tools. Key principles include:

		 Engaging healthcare workers to design and provide feedback on health informatics tools.

		 Promoting electronic error-reporting systems as a vehicle to identify errors and swarm and solve problems in real time.

		 Leveraging technology such as computer screen savers and handheld electronics to present safety data to healthcare workers in the course of their daily work.

		 Educating all healthcare workers in health information technology; developing and promoting a cadre of experts in the field to meet the growing needs of our healthcare industry.

Despite increasing awareness of the risks to patients within the US healthcare system and the human and economic toll that medical errors exact, improvements in patient safety have been slow.

 While individual accountability for patient outcomes is essential for the delivery of high-quality medical care, the complexity of healthcare requires a highly reliable system built within a culture that recognizes errors and process defects as opportunities to learn and continuously improve. The Institute of Medicine report

 emphasized the important role of health information technology in preventing harm to patients.

 Through the automation of errorprone tasks and decision support systems designed to minimize reliance on human memory, health information technology has rapidly become an important tool to address problems faced by healthcare organizations and their patient safety programs.

In this essay, we describe the fundamental role of health information technology as a multifaceted and indispensable tool to achieve high reliability in healthcare. We illustrate how health information technology can contribute to the creation of high performance by discussing the operating characteristics of high-performing organizations and how they achieve superior outcomes. Finally, we caution readers that this technology has the potential to disappoint or even harm if poorly designed, implemented, or embedded within an institution that lacks a culture of safety. We write these opinions in our roles as healthcare leaders in an academic medical center that has a well-developed computer physician order entry system and a partially developed electronic medical record.

The Goal: American Healthcare as Best in Class

, Dr Steven Spear cogently outlines the characteristics of high-performing organizations that produce products that are “best in class.”

 A high-performing organization is an organization which embeds process improvement as a guiding principle and is committed to learning by understanding and solving real problems while developing problem-solving capabilities in all workers. Best-in-class products or solutions are those which are implemented reliably and affordably, and have sustainable outcomes. High-performing organizations achieve their results not through technological advances, but through the complete engagement of the wisdom, knowledge, and skills of every worker. They predict results by implementing a change, measuring some component of the change process, and then compulsively comparing the actual outcome with the anticipated result. Any gap between what was anticipated and what actually occurred is the breeding ground for learning and information dissemination. Such organizations cope with complexity by focusing continuously on how to improve the work that they do.

The culture within high-performing organizations is built upon an organizational learning framework that facilitates and rewards problem-solving capabilities. According to Spear, these capabilities include capturing all existing knowledge, solving problems in real time to build new knowledge, rapid dissemination of knowledge, and building problem-solving and improvement skills in all workers. We will describe how health information technology impacts each of these capabilities below.

Specifying Work to Capture All Existing Knowledge

Problem-solving skills begin with an in-depth understanding of the nature of the current situation. Detailed observations of existing processes by individuals engaged in the work serve as the foundation for work redesign. Increasingly, information technology serves as the repository of existing knowledge in healthcare. In addition to serving as the central location for clinical documentation in many hospitals and health clinics, electronic medical records and computer physician order entry systems (CPOEs) are driving clinical decision making and healthcare delivery at the bedside. We have several examples from our institution to illustrate this concept. In an effort to improve anticoagulation safety and reduce medication dosing errors at our institution, our CPOE system now displays recent anticoagulant doses and lab values for physicians when a blood thinner is prescribed. Information on multi-drug—resistant organisms is captured and displayed prominently in our information systems so that patients who have had a prior resistant infection can be isolated immediately upon entering our emergency department. This reduces the risk of spreading these infections. A new information technology application within the CPOE allows a common legible electronic format for nursing assessments at the bedside. This technology includes vital signs which can be trended so that significant variations are more easily apparent.

In each of these scenarios, reliance on human memory is decreased and the availability of relevant information at the point of care is increased, thereby allowing

physicians to focus on the clinical decision at hand rather than being frustrated and distracted with the process of information gathering.

Conversely, we are challenged when all existing knowledge cannot be captured by information technology, and there are numerous examples of this challenge in healthcare. A prominent example is medication reconciliation. Medication reconciliation is a complex task that requires physicians to compare a patient’s home medication list with their hospital medication list in an effort to identify discrepancies and prevent medication errors. In our medical center, as in many others, this is challenging because more than 1 information system is employed and medication lists from disparate systems are not merged into 1 list or visible on a single computer screen for side-by-side comparison. Pending laboratory tests at hospital discharge are another example of this challenge. Many healthcare systems lack the information

technology to alert discharging and primary care physicians about important tests that are pending at discharge. Promising examples of health informatics solutions related to medication reconciliation and pending test results at discharge have been reported, yet much more work remains to be done in order to achieve high reliability at the intersection of patient safety and health information technology.5,6 These and other similar problems are ideal quality improvement targets for healthcare leaders. Conversations with the healthcare workers who use information technology at the point of patient care is a starting point for healthcare leaders who want to increase the usability of existing knowledge in their systems and learn where and how their informatics tools could be improved.

Swarm and Solve Problems to Build New Knowledge

in Real Time to Avoid Information Atrophy

The primary approach to problem solving in high-performing organizations is to probe deeply into problems to find a rich new source of knowledge. A critical component

of this approach is the examination of identified problems in real time. Information systems play an important role in this process by gathering data and transforming them into action. Perhaps the most visible example of this in the patient safety movement is electronic error-reporting systems.

Prior to electronic error-reporting systems, so-called “incident reports” often lay physically dormant for months in the offices of busy quality improvement directors or nurse managers. It could take weeks to circulate a report to the appropriate person responsible for action. This time delay put healthcare organizations and their patients at risk because the error or “new problem” was not known for a prolonged period, during which time similar incidents could occur.

Electronic error-reporting systems can be readily structured into domains such as what the problem was, when it occurred, where it occurred, who was involved, and ultimately after root cause analysis, why it occurred. This allows repetitive themes (eg, specific units, time of day, day of week, type of medication) to be identified and adds rich context to problem solving. For example, an electronic error-reporting system might identify nights as the most frequent time when a certain class of errors occurs. A specific illustrative example from our hospital’s electronic error-reporting system and review process is a problem of late-evening falls on our heart failure unit as patients arise to urinate. These falls were entered into the reporting system and traced to evening doses of diuretics, which prompted a change in the time of administration.

The ability to “swarm and solve” errors in real time, as described by Spear, was inconceivable for many healthcare leaders prior to the advent of electronic error-reporting systems. Now, armed with the technology that electronic error-reporting systems afford, healthcare leaders and organizations can avoid the inevitable information “atrophy,” speculation, and loss of employee trust that occurs with delayed investigations of errors and their precursor conditions. In our institution, electronic error reports are routed to and read by risk managers as well as the physician, nurse, and quality leader of the specific hospital unit or department where the event occurred.

While risk managers perform the primary investigation for errors with harm, the local leadership leads the investigation of near-miss safety events and trends within their unit or area. Some problems are solved locally, while others that are more complex require collaboration with other leaders in a root cause analysis or related learning forum. This allows “swarm and solve” problem solving to begin locally before memories fade and information atrophies.

Share and Disseminate the New Knowledge Rapidly Throughout the Organization

Perhaps the most powerful application of health information technology related to patient safety is the ability to rapidly disseminate knowledge about error-prone conditions throughout an organization in order to learn, mitigate risk, and improve outcomes. New knowledge about patient safety comes from innumerable sources such as voluntary error-reporting systems, root cause analyses, patient complaints and surveys, and the daily work of clinical staff.

Although e-mail has become the predominant mode of communication in most healthcare organizations, in our experience we have found that the volume of e-mails often leads to important safety information being overlooked or ignored. Additionally, e-mails are frequently read during non-clinical work time, yet the information contained within them would be more useful and effective if it were read in the normal work flow of clinical care. Thus, the development of novel communication methods that present information to physicians, nurses, and other healthcare workers in a “just in time” fashion must be explored.

Since new knowledge about individual patients is often gained through technology portals, it follows that more broad-based knowledge related to safety trends could be disseminated through the same portals. For example, the “idle” computer screen itself can be a portal, and the use of computer screen savers as an educational tool has been described.

 Since most healthcare workers already use personal digital assistants (PDAs), smartphones, or other similar handheld electronic devices to retrieve and organize factual medical information and personal information, it is conceivable that we will use these same devices to communicate safety information rapidly in the near future. Descriptions of healthcare interventions using PDAs have already been described.

 Information overload remains a concern with this mode of communication as well, and strategies to triage safety information and alerts based upon priority and risk scores will be necessary.

Moreover, information systems can be used to share quality improvement and error-reporting data in real time, thereby engaging all healthcare workers in the process and outcomes that they are creating. This requires building systems that can be used and understood by engaged workers and not shrouded in metrics that conceal the human elements of the work. A common taxonomy for patient safety work that includes a standardized vocabulary and definitions is helpful when building such a system. Restricting access to a few individuals or aggregated outcomes only limits information sharing, degrades trust, and creates barriers to healthcare worker engagement. The ability to create interactive and usable websites to display quality scorecards at the unit (micro) and hospital (macro) level should be available to all clinicians.

Lead by Developing These Skills in All Workers

Leaders in high-performing healthcare organizations not only must ensure that their staff have basic informatics skills, but also must identify, mentor, and develop clinical leaders in the burgeoning field of health information technology.

In addition to the training provided during the implementation of new informatics tools, more formal education in the field of health information technology is being called for earlier in the training of healthcare professionals. For example, informatics is now included as a competency for quality and safety education in undergraduate and graduate nursing.

 Similarly, the topic of educating physicians in this field is being discussed and should be included in curricula focusing on quality, safety, and healthcare systems.

 By training all healthcare workers in the availability and use of informatics tools for patient safety, healthcare leaders can leverage these tools as they develop problem-solving capabilities within their workforce, empowering them to identify opportunities for informatics innovation in order to improve patient outcomes.

In parallel, healthcare organizations will need to identify and develop healthcare leaders with interest and skills in medical informatics in order to achieve excellent outcomes. Advanced training certificates, fellowships, and competencies in biomedical informatics exist and will likely expand to meet the growing needs of the healthcare industry.

 In addition, the field of human factors engineering, which evaluates the error-prone properties or “factors” of humans and considers them in the design and deployment of systems, is growing rapidly in order to increase the usability and safety of new and existing health information systems. Indeed, there is an increasing need for human factors engineers in robust biomedical informatics and patient safety programs.

There is a tendency to believe that information technology itself will solve much of what is defective in our healthcare system. This is a dangerous posture, as a multitude of new errors have been introduced into healthcare as a result of technology. 

 We assert that in order for information technology to enhance safety, the organization in which it is embedded must adhere to the fundamental principles of highly reliable organizations: preoccupation with failure, sensitivity to operations, reluctance to simplify, a commitment to resilience, and deference to expertise.

Examples from other industries clearly indicate that a culture of continuous learning and preoccupation with failure trumps technological advantage. This is most evident in the automobile manufacturing industry, where Toyota’s inexorable rise to the top occurred despite similar technology being used by its American and Japanese competitors.

The impact of work flow design on health information technology is often overlooked, and its importance cannot be overemphasized. The implementation of each new informatics or clinical decision support tool gives us the unique opportunity to re-examine how work is done in a detailed fashion and to redesign faulty, redundant, or antiquated processes. Despite this opportunity, many informatics tools continue to be implemented without attention to work flow and with insufficient input from clinical users. For example, in one study of bar code medication administration at 5 hospitals, nurses overrode bar code medication alerts for 4% of patients charted and 10% of medications charted.

 The end result of inattention to work flow and process design is automation of a bad process which leads to user frustration, non-adoption of the tools, workarounds, and perhaps most importantly, a loss of faith that information technology can improve daily work and patient safety outcomes.

Articles

WEB EXCLUSIVE

This commentary reflects how high-performing healthcare organizations use health information technology to advance patient safety.

Chasing High Performance: Best Business Practices for Using Health Information Technology to Advance Patient Safety

To determine the prevalence of physician incentives for quality and to test the hypothesis that the quality of ambulatory medical care is better when provided by physicians with

Cross-sectional study using data from the National Ambulatory Medical Care Survey.

We examined the association between 12 measures of high-quality ambulatory care and physician compensation based on quality, physician compensation based on satisfaction, and public reporting of quality measures.

 Overall, 20.8% of visits were to physicians whose compensation was partially based on quality, 17.7% of visits were to physicians whose compensation was partially based on patient satisfaction, and 10.0% of visits were to physicians who publicly reported quality measures. Quality of ambulatory care varied: weight reduction counseling occurred in 12.0% of preventive care visits by obese patients, whereas no urinalysis in patients with no indication was achieved in 93.0% of preventive care visits. In multivariable analyses, there were no statistically significant associations between compensation for quality and delivery of any of the 12 measures, nor between compensation for satisfaction and 11 of the 12 measures; the exception was body mass index screening in preventive visits (47.8% vs 56.2%, adjusted 

 = .004). There was also no statistically significant association between public reporting and delivery of 11 of 12 measures; the exception was weight reduction counseling for overweight patients (10.0% vs 25.5%, adjusted 

We found no consistent association between incentives for quality and 12 measures of high-quality ambulatory care.

Using data from the National Ambulatory Medical Care Survey, we examined the association between 3 incentives—financial compensation for quality and patient satisfaction, and public reporting of quality measures&mdash;and 12 ambulatory care quality measures.

		 We found fewer than one-fifth of ambulatory visits were to physicians receiving these incentives.

		 There were no consistent associations between financial incentives or public reporting and 12 measures of ambulatory quality.

Variation in quality is a problem of the US healthcare system.

 Pay for performance and public reporting of quality measures are 2 incentives that health insurers and payers use to promote high-quality medical care.

 Almost half of commercial health maintenance organizaions (HMOs) use some form of pay for performance, according to a 2006 national survey.

 In addition, the Deficit Reduction Act of 2005 and the Tax Relief and Health Care Act of 2006 required the Centers for Medicare & Medicaid Services (CMS) to establish plans for value-based payments and physician quality reporting mechanisms.

 In 2007, CMS instituted a voluntary physician reporting program—the Physician Quality Reporting System&mdash;as part of its valuebased purchasing program.

 CMS also has an ongoing demonstration project, the Physician Group Practice project, to study the impact of pay for performance and other payment models on quality of care.

Patient Protection and Affordable Care Act of 2010 has allocated funding for models such as accountable care organizations and the Patient- Centered Medical Home, which tie compensation to quality.

Despite increasing interest, the impact of pay for performance and public reporting on the quality of ambulatory medical care is unclear.

 A 2006 systematic review of the literature found that very few studies had assessed the effect of pay for performance on quality.

 Of these studies, the impact of pay for performance varied: several programs showed improvements in quality and others showed little or none. Even less is known about the effect of physician public reporting of quality measures on ambulatory quality of care.

 Furthermore, evaluations of physician-level incentives to improve quality have focused on a small number of health plans or systems, or programs outside the United States.

 To our knowledge, no studies have examined the association between physician incentives and quality of ambulatory care on a national level.

Using a nationally representative survey of ambulatory visits in the United States, we sought to determine the prevalence of physician incentives for quality and to test the hypothesis that quality of ambulatory medical care is better in the context of these physician incentives. We looked specifically at financial incentives that reward higher quality care and financial incentives that reward better patient satisfaction with the hypothesis that these methods of compensation will be associated with higher quality care. We also examined the association between quality and public reporting of quality measures with the assumption that physicians are motivated to improve performance on publicly reported quality measures to maintain or improve their professionalreputation and patient volume.

We performed a cross-sectional analysis using data from the 2006 and 2007 National Ambulatory Medical Care Survey (NAMCS). NAMCS is a nationally representative survey administered by the Centers for Disease Control’s National Center for Health Statistics. NAMCS contains information about patient visits to non—federally funded, non–hospitalbased offices throughout the United States. Physicians in the fields of anesthesiology, radiology, and pathology are excluded from the survey. Physicians who participate in the survey cannot participate again for at least 3 years.

NAMCS uses a 3-stage sampling design. The first stage is based on geographic location, the second stage identifies offices in each geographic location, and the third stage samples visits within each office. The visits sampled take place during a 1-week period that is randomly assigned for each practice. Between 20% and 100% of the visits that week are sampled, depending on the size of the practice. The National Center for Health Statistics weighs each visit so that the data can be used for national estimates. Each visit weight accounts for selection probability, adjusts for nonresponse, and accounts for other factors so that the national estimates properly reflect the scope of ambulatory visits in the United States.

The survey collects patient and office demographics, and visit-specific clinical information. The information from each visit is recorded on a standardized survey form by the physician, office staff, or a US Census Bureau representative. Clinical characteristics include up to 3 reasons for the visit coded using a reason-for-visit classification, up to 3 diagnoses coded using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), and up to 8 drugs coded using the Lexicon Plus classification, a proprietary database of Cerner Multum, Inc. Because this study used publicly available NAMCS data without respondent identifiers, the Mount Sinai Institutional Review Board exempted it from review.

We limited our sample to visits by nonpregnant patients who were >18 years of age. Because this study focused on rates of delivery of high-quality ambulatory care, we further limited our sample to visits to primary care physicians (ie, general internists, family practitioners, gynecologists) and internal medicine subspecialists (eg, cardiologists, endocrinologists). We excluded visits to surgical specialists and nonmedical specialists (eg, dermatologists, psychiatrists, occupational medicine) because the measures of high-quality care of interest did not pertain to their areas of practice. Finally, we excluded visits in which a physician was not seen.

We hypothesized that 3 physician incentives routinely collected as part of the NAMCS physician induction survey

 would be associated with better quality of care. First, visits were categorized as whether or not the physician’s patient care compensation was at least partially based on quality measures. Second, and similarly, visits were categorized as whether or not the physician’s patient care compensation was at least partially based on patient satisfaction. Of note, the proportion of physician compensation based on quality or satisfaction was reported in quartiles (<25%, 25%-50%, 51%-75%, and 76%-100%). For our study both incentives were categorized as any or none because in the large majority of visits (93.7%), incentives accounted for less than 25% of compensation. Finally, visits were categorized as whether or not the physician or practice publicly reported quality measures.

 We examined 12 measures of ambulatory quality of care (11 process measures and 1 outcome measure) that were identified using the Physician Quality Reporting System measures from CMS (

 The Physician Quality Reporting System measures include 216 processes and outcomes of care, but the majority of these measures could not be examined because the measure (1) was not related to ambulatory care; (2) could not be analyzed on a per visit basis; (3) was not systematically collected in NAMCS; or (4) examined a disease or condition for which the sample of eligible visits was too small (<30 visits) for accurate analyses.

Eligible visits for our study included visits to physicians for preventive care; chronic disease care for 5 conditions (diabetes mellitus, heart failure, coronary artery disease, atrial fibrillation, and chronic obstructive pulmonary disease [COPD]); and upper respiratory tract infections. Visits were identified using 

 codes. Preventive care visits were defined as visits to the patient’s primary care physician with 1 of the following: (1) 

 code V70.0X or V70.9X (or both); (2) general medical examination as a reason for the visit (reasonfor- visit code 31000); or (3) a preventive or nonillness care code as the NAMCS major reason for visit.

We examined 4 measures of high-quality ambulatory care during visits to physicians for preventive care: smoking cessation counseling for smokers, body mass index (BMI) screening, weight reduction counseling for overweight patients, and urinalysis not performed or ordered. Of note, the urinalysis measure was the only measure that we considered achieved if the test was not performed. For this measure we excluded visits by patients with urinary symptoms, renal disease, diabetes, HIV, AIDS, or hypertension.

We examined 1 measure of high-quality diabetes care: blood pressure measurement of less than 130/80 mm Hg.

We examined 2 measures of high-quality heart failure care: (1) prescription of either angiotensin-converting enzyme inhibitor or angiotension receptor blocker therapy, excluding visits by patients with hyperkalemia or angioedema; and (2) prescription of beta-blocker therapy, excluding visits by patients with heart block, bradycardia, COPD, or asthma.

We examined 2 measures of high-quality coronary artery disease care: (1) prescription of oral antiplatelet therapy, excluding visits by patients with peptic ulcer disease, gastritis, gastrointestinal bleeding, duodenitis, or renal disease; and (2) prescription of beta-blocker therapy, excluding visits by patients with heart block, bradycardia, COPD, or asthma.

Finally, we examined 3 additional measures of high-quality ambulatory care: (1) no prescription of antibiotic therapy during visits by patients for upper respiratory infection, excluding visits by patients with COPD, HIV, AIDS, or cancer; (2) prescription of anticoagulation therapy during visits by patients with atrial fibrillation, excluding visits by patients with peptic ulcer disease, gastritis, gastrointestinal bleeding, duodenitis, cerebral hemorrhage, central nervous system tumors, renal disease, thrombocytopenia, or gait abnormality; and (3) prescription of bronchodilator therapy during visits by patients with COPD. For each measure of high-quality care, eligible visits were categorized by whether or not there was documentation that the patient received the recommended ambulatory care.

All visits were additionally categorized by other patient and physician characteristics. For patient characteristics, we examined the following: age, sex, race/ethnicity, number of chronic conditions, and insurance type (private, Medicare, Medicaid, and other, which included self-pay, worker’s compensation, and no fee). For physician and practice characteristics, we examined practice size (solo or group), practice type (private practice, community health center, or HMO), physician employment status (owner, employee, or contractor), physician specialty (primary care or medical specialist), region of practice (Northeast, Midwest, South, or West), and urban or rural practice location.

We performed a visit-level analysis using visit-level sampling weights to account for physician and practice clustering.

 We described visit characteristics using standard frequency analyses, presenting them as the weighted proportion of visits in our study sample. We used the c2 test to study the bivariate association between physician incentives or public reporting and delivery of each of the 12 quality indicators. We used multivariable logistic regression to assess the independent effect of physician incentives on the delivery of each of the 12 quality indicators, creating independent models for each outcome while controlling for the patient and physician characteristics as outlined above and whether the physician received other incentives for quality or publicly reported quality measures. We also performed logistic regression to determine whether more incentives or specific combinations of incentives had incremental associations with quality.

In order to understand whether the results of our study differed among subgroups of patients, we performed additional sensitivity analyses stratifying patients by insurance type (private insurance, Medicare, and Medicaid) and age (<65 and 

All analyses took into account the complex survey design and weighted sampling probabilities of the data source for the calculation of nationally representative point and variance estimates. All analyses were performed using Stata statistical software, version 11.0 (Stata Corp, College Station, Texas). Because we repeated our analyses for 3 different physician incentives, we used the Bonferroni correction and used a P value of .01 to signify statistical significance.

Sociodemographic and Clinical Characteristics

We found no consistent associations between physician incentives for quality improvement and 12 measures of ambulatory quality of care.

Association Between Physician Quality Improvement Incentives and Ambulatory Quality Measures

 To examine patient and physician factors affecting utilization of diagnostic imaging in primary care.

 Patient-level data from a large academic group practice over the period July 1, 2007, through June 30, 2009.

This is a retrospective cohort study of 85,277 patients cared for by 148 primary care physicians (PCPs). The dependent variable is the number of outpatient imaging exams ordered by each patient’s PCP over the study period. Independent variables include 17 patient factorsdescribing both clinical need and demographic characteristics and 7 physician factors.

Data were collected from the electronic medical record and associated administrative databases.

Patient factors having a statistically significant effect on both the probability race, more than 10 medications, congestive heart failure, diabetes, hypertension, other problems, visits to the PCP, visits to specialists, and imaging exams ordered by specialists. For physician factors, experience, gender, and having another degree were statistically significant in both portions of the model.

Both patient and physician factors have a substantial effect on primary care outpatient diagnostic imaging utilization. Several of these significantly influence both the probability that any images will be ordered and the intensity (number) of imaging.

		 Primary care providers caring for adult patients order from 0 (63%) to 15 images per patient over 2 years, with the average being 0.72.

		 Multivariable modeling of outpatient imaging utilization should take into account the observed distribution of per patient counts, which is 0 inflated and skewed to the right.

		 Nine of 17 patient factors influence both the probability of any imaging and the amount of imaging performed.

		 Three of 7 provider factors influence both the probability of any imaging and the amount of imaging performed.

A central problem facing health policy makers is how to control the costs of medical services while maintaining the quality of care. Given that costs depend on utilization and unit costs, an understanding of factors driving utilization is fundamental to developing feasible and effective recommendations for cost control. Diagnostic imaging utilization is of particular interest because medical technology is often mentioned as a major driver of costs, with radiology in particular experiencing rapid growth. For example, in their 2006 report to Congress, MedPAC noted that from 1999 through 2004 growth in Medicare claims for diagnostic imaging was the highest of all physician services, at 62%.

Much of the research on utilization of health services stems from the large body of work that has demonstrated considerable variation in their use at various levels of geographic comparison.

 Much of this variation in the use of medical services is not explained by differences in health status, disease-specific population health indicators, or other measures of health outcomes. This raises the question of what other factors influence utilization.

Although there is a large literature on the general topic of utilization of healthcare services, less attention has been paid to diagnostic imaging. Some studies focused on diagnostic services in general

 examined utilization of outpatient diagnostic imaging and found that the probability of a diagnostic imaging study being ordered was significantly affected by patient age, urgent visits, visit frequency, and physician gender. A study by Grytten and Sorensen

 examined the use of diagnostic tests (including imaging) among Norwegian primary care physicians (PCPs); after controlling for patient age, gender, and reason for visit, the remaining variation in diagnostic test use ranged from 47% to 66%. Other studies have examined computed tomography (CT) scans and magnetic resonance imaging (MRI) only.

The objective of this study is to examine patient and physician factors determining utilization of diagnostic imaging in primary care. Detailed patient-level data come from the physician organization associated with a large, urban academic medical center over a 2-year period. A patient-provider connectedness algorithm identifies a cohort of patients each linked to a single PCP based on outpatient visits over the previous 3 years. The analysis uses a 2-part model, which distinguishes between the probability of any imaging utilization (yes or no) and, conditional on any imaging, the amount of imaging utilization (number of exams). Multivariate regression is performed using maximum likelihood methods and zero-inflated Poisson distribution, which allows simultaneous estimation of coefficients for the 2-part model.

The study contributes to the literature by providing a deeper understanding of factors that influence the utilization of diagnostic imaging in primary care. The data used in the study are rich in detail and depth; the analysis includes 17 patient factors describing both clinical need and demographic characteristics as well as 7 physician factors, including gender, experience (years after MD degree), panel size, and having been sued for malpractice. Also, previous research on imaging utilization typically has focused solely on the amount of utilization, whereas this study applies the 2-part model in order to distinguish between the probability of any utilization and the amount of utilization, and the methodology used jointly estimates coefficients for the 2-part model.

This study examines the utilization of outpatient diagnostic imaging by PCPs caring for adult patients. We will not discuss imaging that occurs as part of disease screening programs, such as mammography and CT colonography, because the factors affecting utilization are likely to differ between imaging for screening and imaging for diagnostic purposes. Factors influencing the utilization of outpatient diagnostic imaging can be divided into 4 major classes: patient, physician, organization, and community. Because this study uses data from a single large physician organization in a single community, the focus is limited to patient and physician factors.

Patient factors affecting imaging utilization stem from clinical need as modified by demographic and socioeconomic characteristics. The physician orders imaging to address clinical uncertainty that arises about an existing medical condition or clinical event. Under ideal conditions of evidence-based practice, the mixture and amount of imaging that patients received would depend only on their existing diagnoses, disease status, and clinical events. However, the current state of medical knowledge and the consistency with which it is applied in actual practice is such that physicians’ decisions about intensity and mixture of diagnostic imaging are quite variable even when faced with identical clinical scenarios. Consequently, utilization decisions also are influenced by patient factors other than clinical need as well as by physician factors.

Patient demographics such as gender and age strongly correlate with the amount and type of clinical need based on their complex relationships with many diseases and health states. However, these same demographic factors also exert social and psychological effects on the patient’s likelihood to undergo diagnostic testing under various scenarios. Other patient factors such as ethnicity, level of education, and income also may affect individual patient tendencies to seek care and comply with provider recommendations. In addition to patient factors, physician characteristics may affect a doctor’s tendency to order diagnostic imaging tests. Physician training has direct effects on test-ordering behavior in primary care. Because the majority of training and experience about radiology is gained during residency, the relative intensity of imaging utilization at the training institution strongly influences subsequent decision making about imaging during practice. Previous research has found that physicians trained in high-use regions and at large academic centers tend to order diagnostic imaging more frequently than those trained in smaller centers and low-use regions.

 Physician experience is also relevant, with empirical evidence indicating that physicians having greater experience tend to order fewer imaging tests in identical scenarios.

Physician gender also may have an effect on use of imaging independent of experience and training, although, with the recent and substantial increase in the fraction of women trained in and practicing medicine in the United States, it may be difficult to measure, since women physicians tend to be younger and trained later. Empirical evidence about physician gender has been mixed, with studies showing both increased and decreased tendency to order imaging and other diagnostic tests between male and female doctors in primary care.

Malpractice deserves special mention because the term “defensive medicine” is often used to describe the phenomenon of physicians’ increased ordering of imaging and other diagnostic tests based on fear of being sued for failure to diagnose.

Being sued for malpractice is an extremely negative and unsettling event that induces a strong desire to avoid repeat occurrences.

 Thus, if physicians have been sued for malpractice, the expectation is that they will alter future practice toward more diagnostic testing.

For purposes of the empirical analysis, the conceptual framework can be summarized as: imaging utilization = f [patient factors (clinical need, demographic characteristics), physician factors].

This study was conducted at Massachusetts General Hospital (MGH) and the associated Mass General Physician Organization (MGPO). The Institutional Review Board (IRB) and privacy office approved the study and deemed it to be Health Insurance Portability and Accountability Act compliant. Within the MGPO, the Primary Care Operations Improvement Group conducts analyses of the practice for internal quality assessment and improvement. In support of these efforts they have devised and validated a method for identifying a group of patients and PCPs with stable relationships to each other. The result is termed a connectedness (“loyalty”) cohort. Each cohort is identified by year and comprises a list of patients who are considered to be linked to a single PCP by virtue of their outpatient visits as documented in the electronic medical record (EMR) over the 3 years ending in the “cohort” year. All patients seen in the primary care practice are initially eligible for inclusion in the cohort. Assertions about connectedness for patientdoctor pairs are calculated probabilistically from office visit history.

 During the validation studies of the algorithm, cutoff probabilities were selected to produce sensitivity of 80.4%, specificity of 93.7%, and positive predictive value of 96.5%. The original goal of the method was to achieve a positive predictive value of 90% for any given provider’s linked patients.

The Radiology Department at MGH provides a full range of imaging services for inpatient, emergency department, and outpatient practices. The main department is located at the MGH campus, with several ancillary outpatient sites in the Boston, Massachusetts, area. The whole department is linked via a robust electronic infrastructure with records of all imaging tests housed in a data warehouse. In total, almost 100 items of information are stored about each test, including the identity of the ordering physician, dates of order and completion, modality, body area, and setting.

The 2008 primary care connectedness cohort contained 87,568 patients flagged as being linked to a PCP. There were 804 patients who were linked to 26 providers with less than 100 patients in their practices. These were excluded, leaving 86,764 patients. Of these, 1483 were linked to 4 providers who had left the MGPO in late 2008 or the first or second quarters of 2009. Therefore the analytic sample includes 85,277 patients, linked to 1 of 148 PCPs. These physicians practice in 1 of 15 clinics distributed through the greater Boston area. The unit of observation for this study is the patient. The 85,277 patient medical record numbers were queried against the clinical radiology system to return all diagnostic imaging exams performed during the study interval (July 1, 2007, through June 30, 2009). The query specifically excluded interventional procedures and mammograms. Table 1 shows the total number of outpatient diagnostic imaging exams performed on the study cohort during the study period. The dependent variable for the study was constructed by aggregating and summing (by patient) the 60,983 outpatient imaging procedures ordered by the PCP to whom the patients were linked (outpatient ordered by PCP in Table 1). The remaining diagnostic imaging procedures were performed while patients were in the emergency department (N = 34,345), were completed while patients were in the hospital (N = 29,763), or were outpatient exams ordered by specialists (N = 96,525). These other categories of imaging utilization were also aggregated by patient and summed to produce 3 patient-level imaging utilization independent variables as described below.

The EMR system was used to obtain counts of various clinical events for each patient over the study period. Hospital activity variables included visits to the emergency department (ED), inpatient hospital stays, intensive care unit (ICU) days, and inpatient observation days. Records of outpatient visits for each patient during the study period were used to sum the professional relative value units (RVUs; obtained from Centers for Medicare & Medicaid Services as of 2008) for visits to the linked PCP and to specialists.

The EMR system was used to develop an active medical problem list for each patient. The coded problems were mapped into the following broad categories: diabetes, hypertension, heart failure, coronary artery disease, renal failure, cancer, trauma, obesity, and substance abuse. For each of the major categories, a binary variable was constructed when the patient had at least 1 active problem listed falling into that category. The counts of unmapped problem codes were placed into a separate variable of other problems.

The EMR system also provided data on the number of outpatient medications each patient was taking during 2008. The queries did not count refills of the same drug and dose as new prescriptions. However, switches within a drug class and/or dose changes were counted as new prescriptions, which could result in over-counting and rendering small differences in the discrete number less meaningful than the general amount each patient was taking. Therefore, we stratified the count of outpatient medications into 4 categories.

The MGH registrars’ database was used to obtain relevant information pertaining to the 148 PCPs in the study, including gender, birth year, medical school graduation year, and medical school state (or country for foreign medical graduates). All were licensed in the state of Massachusetts, and the publicly available website for the Massachusetts Board of Medicine was queried to determine whether or not each physician had been sued for medical malpractice in the past 10 years.

We examine and quantify patient and physician factors that affect the utilization of outpatient imaging by primary care physicians caring for adult patients.

Determinants of Diagnostic Imaging Utilization in Primary Care

 There are over 12 million ambulatory care visits for acute pharyngitis annually in the United States. Current guidelines recommend diagnosis through culture or rapid antigen detection test (RADT) and relatively straightforward treatment. Community pharmacists may provide cost-effective care for disease states such as group A streptococcus (GAS) pharyngitis.

 The objective of this research is to evaluate the cost-effectiveness of a community pharmacist—as-provider program for the diagnosis and treatment of pharyngitis caused by GAS as compared with standard of care.

A cost-effectiveness analysis was conducted to compare treatment for adult pharyngitis patients. In addition to 5 physician-provided treatment strategies, the episodic costs and benefits of treatment provided by pharmacists using RADT and walk-in clinics using RADT were also considered. Model parameters were derived through a comprehensive review of literature and from the Centers for Medicare & Medicaid Services physician fee schedule. Utilities were expressed in quality-adjusted life-days (QALDs) to account for the relatively short duration of most cases of pharyngitis.

 Using a cost-effectiveness threshold of $137 per QALD, GAS treatment provided by a pharmacist was the most cost-effective treatment. Pharmacist treatment dominated all of the other methods except physician culture and physician RADT with follow-up culture. The incremental cost-effectiveness ratio (ICER) for physician culture was $6042 per QALD gained and $40,745 for physician RADT with follow-up culture.

This model suggests that pharmacists may be able to provide a cost-effective alternative for the treatment of pharyngitis caused by GAS in adult patients.

This study is the first to show that community pharmacy—based rapid diagnostic testing and treatment for group A streptococcus (GAS) pharyngitis may be a cost-effective treatment strategy.

		 Improved rapid diagnostic testing and well-known treatment protocols make GAS an ideal disease for treatment in alternate settings including community pharmacies.

		 Treatment in a community pharmacy allows for more convenient and less expensive treatment for GAS.

		Policy makers and payers looking to reduce overall healthcare costs should consider supporting cost-effective strategies such as utilizing community pharmacies as primary care providers.

There are over 12 million ambulatory care visits for acute pharyngitis annually in the United States.

 Approximately 15% to 30% of these cases are attributable to group A streptococcus (GAS) in children, and 5% to 10% in adults.

 Treatment of pharyngitis caused by GAS, or strep throat, has traditionally been initiated in the setting of a clinic and performed by a physician or mid-level practitioner. Protocols for the treatment of GAS pharyngitis are available, but there is some debate over which approach is optimal. The current guidelines from the Infectious Diseases Society of America recommend that adults be diagnosed through culture or rapid antigen detection test (RADT) if the physician cannot confidently exclude the diagnosis of GAS on epidemiological or clinical grounds. The Infectious Diseases Society of America guidelines do not recommend confirmation of a negative RADT test with culture in adults.

 In actual practice, physicians are likely to use a variety of strategies when diagnosing and managing GAS pharyngitis. These include observation with no testing or treatment, empiric treatment of all patients with suspected GAS pharyngitis, testing with culture alone, testing with RADT alone, and testing with RADT with a follow-up culture for negative results.

In recent years, pharmacists in the community setting have taken on more diversified roles as frontline healthcare providers. Examples of these expanded roles include providing vaccinations, offering travel medicine services, managing anticoagulation clinics, and identifying and treating patients with influenza. Another possible expansion of pharmacist services would be the diagnosis and treatment of GAS pharyngitis. While the pharmacist could never replace the physician in the diagnosis and treatment of complex patients, it is possible that pharmacists could provide a cost-effective treatment alternative for disease states such as GAS pharyngitis. For diseases such as GAS pharyngitis, use of a pharmacist- as-provider care model might yield cost savings through reduced provider expenses and improved access to care resulting in fewer missed days of school and/or work and containment of disease.

Few data evaluating pharmacist-provided care for GAS have been published. A study from 1978 concluded that cost savings could be realized through pharmacist management of streptococcal throat infections; however, this study was relatively small and not reflective of contemporary practice.

 In 2002 the National Community Pharmacists Association, in collaboration with US Wellness Inc, implemented a program to train pharmacists to conduct rapid strep testing in the community pharmacy and charge for this service.

 The status of this initiative is unknown, but testing for GAS is not yet a common practice in community pharmacies. Additional studies have evaluated the cost associated with the management of GAS, but these do not consider the possibility of pharmacist-provided care.

This research will provide the first cost-effectiveness evaluation of a pharmacist-as-provider program for the diagnosis and treatment of GAS pharyngitis as compared with standard of care provided by physicians and nurse practitioners.

In order to evaluate the cost-effectiveness of pharmacistprovided care for GAS pharyngitis, a decision model was constructed. This model was largely based on the decision model used in the Neuner study, which consisted of 5 arms designed to determine the best method of treatment of pharyngitis in a physician’s office.

 As with the Neuner model, patients included in this model were adults presenting with acute pharyngitis. High-risk patients, such as the immunocompromised, were not included in our model since those patients generally require additional follow-up to monitor their therapy. Our final treatment model consisted of 7 arms: 1) physician observation only; 2) physician treatment of all patients empirically with antibiotics; 3) physician performing a throat culture on all patients; 4) physician performing an RADT on all patients, and performing a throat culture on those with a negative RADT result; 5) physician performing only an RADT; 6) a nurse practitioner in a walk-in clinic performing an RADT on all patients; 7) a pharmacist performing an RADT on all patients in a community pharmacy setting. The first 5 arms are identical to those of the Neuner study. The nurse practitioner arm was added to assess the effect of a recent rise in urgent care clinics, which are often staffed by nurse practitioners. Although the Neuner study concluded that culturing all patients is the most cost-effective method of diagnosis and management of patients with pharyngitis, we elected to include all 5 arms in our model because this allowed us to compare a pharmacy-based treatment model with various medical practices, not simply what is considered the ideal.

Possible complications of GAS were considered in building the decision tree. These complications included acute rheumatic fever (ARF), peritonsillar abscess, and allergic reaction to penicillin. Other complications reported in the literature but not included in our model include glomerulonephritis, otitis media, and sinusitis. A recent Cochrane review described the effect of antibiotics in preventing glomerulonephritis, otitis media, and sinusitis as compared with placebo.

 For these complications, the difference between the rate of complications with and without antibiotics was determined not to be clinically significant and left the authors to speculate if GAS truly causes these complications or if they are simply associated with one another.

The decision tree model was constructed and analyzed using the TreeAge Pro Suite 2009 version 1.0.2. The 

represents the pharmacy arm of the tree. The entire tree can be seen in Appendices A-H.

As with the Neuner model, certain assumptions were made. First, patients were assumed to have no history of ARF or a penicillin allergy. Screening for high-risk patients was assumed to be 100% effective in all settings. Patients developing ARF were assumed to develop no other complication. Patient adherence to therapy is assumed to be 100%. A willingness-to-pay parameter of $137 per quality-adjusted life-day (QALD) was used. This number was based on the common standard of a willingness-to-pay of $50,000 per quality-adjusted life-year (QALY), which was then divided by 365 days in the year.

Our model assumes that the visits in all arms include similar procedures with the exception of the previously described differences in treatment strategy. This includes a basic physical examination of the respiratory system and neck, the collection of vital signs, and detailed written record of the encounter.

It was assumed that pharmacists would be able to carry out the physical assessment and RADT testing with same effectiveness as a physician or a nurse practitioner. It was surmised that pharmacists would be able to dispense antibiotics to the appropriate patients at the pharmacy immediately after RADT testing. This could be possible through a collaborative practice agreement set up with a physician, similar to how collaborative practice agreements are set up for the administration of vaccines in the community pharmacy setting. Finally, it is assumed that there is no cost associated with the establishment of a collaborative practice agreement.

The costs and probabilities used in the model are described in 

. The probabilities used in our study were largely those used in the Neuner study as well as in the Cochrane review.

The original sources of the probabilities were reviewed to verify the quality of the information. The authors of the Neuner study averaged the results of several studies to determine the sensitivity and specificity of the RADT test. We used these averaged values, along with the prevalence of GAS, to calculate the following probabilities: RADT positive test result, true positive, true negative, false positive, and false negative.

Cost information was taken from several sources. The cost of penicillin and treatment of the adverse events of penicillin- induced rash, anaphylaxis, peritonsillar abscess, and acute rheumatic fever were taken from the Neuner study.

 Since these values were given in US dollars from 2000, it was necessary to update these costs to reflect the effect of inflation. From April 2000 to April 2010, the healthcare inflation rate was 1.5, meaning that $1.00 in 2000 would have the same purchasing power as $1.50 in 2010.

 Therefore, the cost of penicillin and the cost of treatment of adverse effects from the Neuner article were multiplied by 1.5 to reflect current costs. The cost of a physician visit was based on the 2010 physician fee schedule from the Centers for Medicare & Medicaid Services (CMS) using a Healthcare Common Procedure Coding System (HCPCS) code of 99213, which allows 15 minutes of face-to-face time with the patient and/or family.

 Cost for a walk-in clinic was also calculated using the CMS physician fee schedule based on the fact that a nurse practitioner is reimbursed at a rate of 85% of a physician.

 Pharmacist reimbursement was based on reimbursement rates for provision of medication therapy management (MTM) services, which are currently billable services to Medicare.

 CMS reimbursement rates for a throat culture and RADT were also used.

 A Current Procedural Terminology (CPT) code of 87880QW was used for the RADT, and a CPT code of 87070 was used for the culture. One should note that practitioner time is considered a separate cost from testing. For example, if a patient went into a walk-in clinic and received an RADT test, the cost would be $55.80 (practitioner time) + $17.18 (cost of test).

Utility information was the same as what was used in the Neuner study, which was given in lost QALDs. A utility of 0.95 per day was assigned to pharyngitis, which is the utility associated with other minor symptoms such as diarrhea and dyspepsia.

 Therefore, the QALDs lost with pharyngitis are based on the duration of symptoms. Untreated pharyngitis is associated with symptom duration of 5 days for a total QALD loss of 0.25. Similarly, pharyngitis treated after culture diagnosis is associated with a symptom duration of 4 days due to the delay in initiating therapy, and pharyngitis treated empirically or after RADT is associated with a symptom duration of 3 days. The Neuner study assigned utilities to anaphylaxis, penicillin-induced rash, and ARF based on a patient survey,

 and assigned utility to peritonsillar abscess based on a previous decision analysis that used expert opionion.

Using our baseline probabilities, costs, and utilities, treatment of GAS pharyngitis by a pharmacist was the most costeffective treatment strategy (measured in dollars/QALDs gained). Pharmacist treatment dominated all of the other methods except physician culture and physician RADT with follow-up culture (

). The incremental cost-effectiveness ratio (ICER) for physician culture was $6042 per QALD gained and $40,745 for physician RADT with followup culture. Both values greatly exceed the $137 per QALD threshold.

As seen in Table 2, the QALDs lost with all of the treatments were quite comparable, with the exception of empirical treatment by a physician. There is a difference of 0.0096 QALDs between physician culture and physician observation. Since the differences among strategies were negligible, this indicated that all treatments (excluding physician empirical therapy) are equally effective. Therefore, a cost-minimization approach may be taken in analyzing the data rather than a cost-effectiveness approach, choosing the treatment method with the lowest overall cost.

A one-way sensitivity analysis was performed on all probability and cost variables using a cost-minimization strategy. The variable for RADT negative GAS positive (false negative) probability was the only variable that changed the study results in the sensitivity analysis.

This cost-effectiveness analysis of care options for group A streptococcus patients is the first to include community pharmacy as a possible point of care.

Cost-Effectiveness of Pharmacist-Provided Treatment of Adult Pharyngitis

 The identification of patients most likely to benefit from care management programs—case finding&mdash;is a crucial determinant of their effectiveness regarding improved health outcomes and reduced costs. Until now, research has mainly focused on claims data—based case finding. This study aimed to explore how primary care physicians (PCPs) select patients for practicebased care management and how risk prediction may complement their case finding.

 We performed 12 semi-structured interviews with PCPs from 10 small- to middle-sized primary care practices in Germany. The interviews focused on their criteria for selecting patients as potential participants in an on-site care management program and how PCPs evaluate claims data—based risk prediction as a case-finding tool. All interviews were transcribed verbatim. We performed qualitative content analysis using the ATLAS.ti software.

 Three major categories emerged from the physicians interviewed: 1) the physicians’ interpretation of the program’s eligibility criteria, 2) physician-related criteria, and 3) patient-related criteria. The physician-related criteria included “sympathy/aversion” and “knowing the patient.” Patient-related criteria concerned care sensitivity in terms of “willingness to participate,” “ability to participate (eg, sufficient language skills, cognitive status),” and “manageable care needs.” PCPs believed that their case finding could be supported by additional information from claims data—based risk prediction.

Case finding for care management programs in primary care may benefit from a structured approach combining clinical judgment by PCPs and claims data—based risk modeling. However, further research is needed to identify the optimal case-finding strategy for practicebased care management.

Primary care physicians refer patients to practice-based care management programs. Until now, there has been little evidence about the criteria they use to select patients and how they experience predictive modeling—based case selection.

		 Primary care physicians select patients whom they rate to be willing and able to participate in a care management program that mitigates their complex clinical needs.

		 Predictive modeling—based case selection is found to be a valuable approach to support case finding for practice-based care management.

For decades, numerous care management (CM) programs have evolved in the United States and elsewhere aimed at improving quality of care and/or reducing healthcare costs. Despite these CM programs’ diversity in terms of intervention targets and designs, the identification of patients most likely to benefit—commonly called case finding&mdash;is essential for both their effectiveness and cost efficiency.

 Therefore, particular attention has been focused on the development and evaluation of practical tools to screen patient populations for potentially eligible participants.

 Because a relatively small percentage of the population generates a disproportionate share of healthcare needs and healthcare spending,

 these tools mainly focus on the identification of current or future high-cost individuals.

Screening questionnaires like “Probability of Repeated Admissions” (Pra and Pra Plus)

 or “Identification of Seniors at Risk” (ISAR)

 have been empirically tested as case-finding tools and have been proved to identify high-risk individuals with manageable health needs. However, their use is limited by survey response rates, potential bias arising from self-report, and respondent burden.

Since the 1980s, several “predictive models” have been developed to predict future healthcare utilization and costs.

 These models are based on data about diagnostic information, prior healthcare utilization, and costs, which can be obtained from health insurance claims data. Predictive modeling (PM) can be used to identify individuals at high risk of future healthcare utilization as potential participants in CM programs.

However, concerns have risen about which identified potential participants may actually participate in and benefit from these programs. A significant proportion of high-risk patients could not be enrolled in CMinterventions due to being unable or unwilling to participate.

Apart from screening questionnaires and PM, physician referral is another case-finding approach. Compared with PMs, primary care physicians (PCPs) select patients at lower risk of future healthcare utilization but with higher probability of actually participating in CM programs.

 Combining both approaches can help to identify patients at high risk of future healthcare utilization who are willing and able to participate in CM programs. Until now, there has been little evidence about the criteria PCPs use to identify potential CM participants. Eligibility criteria for referral may differ between vendor-supported and on-site primary care practice—based CM programs.

Most recently, the concept of accountable care organizations (ACOs) has evolved in order to accelerate Medicare’s 3-part aim of better care, better health, and slower growth in costs.

Criteria used by primary care physicians to select patients for practice-based care management programs were explored in a qualitative study.

April 2012

Chasing High Performance: Best Business Practices for Using Health Information Technology to Advance Patient Safety

Association Between Physician Quality Improvement Incentives and Ambulatory Quality Measures

Determinants of Diagnostic Imaging Utilization in Primary Care

Cost-Effectiveness of Pharmacist-Provided Treatment of Adult Pharyngitis

Primary Care Physicians' Experiences With Case Finding for Practice-Based Care Management

Factors of Hyperlipidemia Medication Adherence in a Nationwide Health Plan

Predictors of Physician Use of Inpatient Electronic Health Records

Evaluation of Human Immunodeficiency Virus and Hepatitis C Telemedicine Clinics

Predicting Costs With Diabetes Complications Severity Index in Claims Data