Steven D. Pearson

Critical Pathways as a Strategy for Improving Care: Problems and Potential

In recent years, intense pressures to reduce the costs of health care have led many health care organizations to seek strategies that reduce resource utilization while maintaining the quality of care [1-5]. Among the most popular of the methods intended to meet this challenge are critical pathways. Critical pathways are management plans that display goals for patients and provide the corresponding ideal sequence and timing of staff actions to achieve those goals with optimal efficiency [6-8]. Interest in critical pathways has increased tremendously during the past several years as early anecdotal reports of their cost-saving potential have been disseminated, usually outside the peer-reviewed medical literature [7, 9, 10]. The rapid push for critical pathway implementation comes from intense competitive pressures and the persistent evidence of unexplained variation in medical practice [11, 12]. Many managed care organizations have added their weight to this process by mandating certain critical pathways or seeking partner hospitals that are willing to develop their own [7]. However, no controlled study has shown a critical pathway to reduce the duration of hospital stay or to decrease resource use, nor has any study shown critical pathways to improve patient satisfaction or outcomes [13]. Nevertheless, like other promising medical technologies, critical pathways are being disseminated before controlled trials have been done to evaluate their effectiveness. Despite the lack of data, an increasing number of physicians will be asked to participate in critical pathway development. Even more will find that their hospitalized patients are already on pathways that they may or may not have endorsed. To enhance the effectiveness of critical pathwaysand minimize the disruption to the patientphysician relationshipphysicians and other caregivers must understand the origin, potential benefits, and potential pitfalls of this new method. Critical Pathways: A New Form of Clinical Guideline Critical pathways have varying formats and are known by many names, including critical paths, clinical pathways, and care paths. Interpreted formally, a critical pathway is the sequence of events in a process that takes the greatest length of time. Like the techniques of continuous quality improvement, critical pathway techniques were first developed for use in industry as a tool to identify and manage the rate-limiting steps in production processes [14-17]. First developed in the 1950s, the Critical Path Method was frequently linked with a similar approach, the Program Evaluation and Review Technique, to coordinate multiple contractors or persons in a project by identifying the key sequence of events, or critical path, the requirements of which would drive the timeline of the overall project [18, 19]. Critical pathway techniques have subsequently been applied to projects as diverse as construction, civil engineering, town planning, marketing, ship building, product design, and equipment installation [6]. Critical pathways were first developed and applied to health care in the 1980s, when prospective payment systems focused greater interest on potential methods to improve hospital efficiency [6]. Most of the first critical pathways in hospitals were developed by nurses for nursing care alone [20, 21], but multidisciplinary teams soon began developing pathways to encompass all aspects of care for hospitalized patients [22-24]. In general, efforts to develop critical pathways in health care have not incorporated the formal techniques used by industrial predecessors to identify the true critical pathway in any care process [18, 25]. Instead, when critical pathways are used to plan medical care, the specific goals usually include the following: 1. Selecting a best practice when practice styles vary unnecessarily. 2. Defining standards for the expected duration of hospital stay and for the use of tests and treatments. 3. Examining the interrelations among the different steps in the care process to find ways to coordinate or decrease the time spent in the rate-limiting steps. 4. Giving all hospital staff a common game plan from which to view and understand their various roles in the overall care process. 5. Providing a framework for collecting data on the care process so that providers can learn how often and why patients do not follow an expected course during their hospitalization. 6. Decreasing nursing and physician documentation burdens. 7. Improving patient satisfaction with care by educating patients and their families about the plan of care and involving them more fully in its implementation. The general format of critical pathway guidelines is the Gantt chart, which outlines the suggested patient care process based on a time-task matrix, listing the components of care in one column and cross-aligning these entries with columns pertaining to time [8]. Figure 1 is an example of such a chart for a critical pathway for patients who have had coronary artery bypass graft surgery. Categories of multidisciplinary staff actions are listed in the first column of the pathway, with specific actions for each day of hospitalization. As indicated in Figure 1, a patients diet is expected to progress successfully from ice chips to clear liquids on the first day after surgery. For all other categories of patient care, critical pathways likewise explicitly mark the transition points of patient progress and lay out a coordinated map of staff activities to achieve those transitions in the most efficient way possible. Figure 1. The first 2 days of a simplified critical pathway for patients who have had cardiac surgery. Critical pathways differ from most clinical guidelines, protocols, and algorithms in several key respects. First, clinical guidelines often address the appropriateness of care by delineating the indications for tests or treatments. Critical pathways, on the other hand, have almost always focused on the quality and efficiency of care after decisions have already been made to admit the patient or perform the procedure. Another way in which critical pathways differ from most clinical guidelines is that they are multidisciplinary in their development and in the scope of their implementation. Critical pathways are also designed along specific timelines, sometimes even in hour-by-hour detail, for indicated actions, and pathways not only spell out these specific actions but also enumerate expected intermediate patient outcomes that serve as checkpoints for the performance of both the patient and the pathway. Yet another distinguishing feature of critical pathways is that their comprehensive design allows them to be used as a part of the patient record, often replacing other documentation entirely [6, 24]. All staff interventions and intermediate clinical outcomes that occur as expected can be simply initialed on the critical pathway document. If staff actions or intermediate patient outcomes do not occur as expected, however, a variance from the pathway is said to have occurred [26]. Variances, too, can be noted on the document, along with an explanation of their cause or causes, and, if needed, a plan can be described to return the patient to the expected course of treatment and outcomes. Critical Pathway Development Topic Selection Critical pathways are typically developed for the hospital care associated with high-volume, high-cost diagnoses and procedures, particularly those for which inefficient variation in the process of care is thought to exist [6]. Surgical procedures, such as coronary artery bypass graft surgery and total hip replacement, lend themselves particularly well to critical pathways because the care process differs relatively little from patient to patient. For this same reason, obstetric procedures such as normal vaginal delivery and cesarean section have also been subjects of pathways in many institutions [24]. For most medical diagnoses, however, patient care has proved more difficult to translate successfully into critical pathways because of the greater heterogeneity among patients and problems [6, 27]. Some institutions have reported that pathways fail when used for medical patients who have either multiple problems and therefore multiple relevant pathways or a problem that does not fit neatly within any single standardized pathway [27-29]. Despite these concerns, however, pathways have been designed and implemented at many institutions for medical diagnoses such as myocardial infarction, stroke, and deep venous thrombosis [23]. Team Composition The group that is organized to develop a critical pathway should be multidisciplinary in order to bring to the table the knowledge and perspectives that are necessary to view the care process in its entirety. Although many institutions have appointed nurses as the leaders of critical pathway teams [6], we have found that having a physician-expert lead each team lends credibility to the pathways and builds a foundation of support among all clinicians. Each pathway team should also have a group facilitator from the hospital administration, a housestaff physician, a member of the quality management department who has expertise in critical pathway methods, and a community-based primary care physician, whose perspective on inpatient care is likely to differ from that of hospital-based physicians. The lack of active involvement by physician-experts is cited as a key reason for the failure of pathway programs; critical pathways that are developed without physician input have ended up sequestered in a part of the medical record where physicians do not often look [27, 30]. Simply gathering physicians, nurses, and other staff around the same table, however, may not be enough to generate the level of teamwork and communication necessary for success. It is an important challenge, especially for physicians trained in an individualistic ethic, to learn how to participate in and lead these teams eff

Patients' Trust in Physicians: Many Theories, Few Measures, and Little Data

Trust is one of the central features of patient-physician relationships. Rapid changes in the health care system are feared by many to be threatening patients’ trust in their physicians. Yet, despite its acknowledged importance and potential fragility, rigorous efforts to conceptualize and measure patient trust have been relatively few. This article presents a synopsis of theories about patient trust and the evolution of methods to measure it. Clinicians, educators, and researchers interested in this area may find this information useful in practice and teaching. The gaps identified in our knowledge about trust can help target new efforts to strengthen the methodological basis of work to understand this vital element of medical relationships.

Active Surveillance Compared With Initial Treatment for Men With Low-Risk Prostate Cancer: A Decision Analysis

CONTEXT In the United States, 192,000 men were diagnosed as having prostate cancer in 2009, the majority with low-risk, clinically localized disease. Treatment of these cancers is associated with substantial morbidity. Active surveillance is an alternative to initial treatment, but long-term outcomes and effect on quality of life have not been well characterized. OBJECTIVE To examine the quality-of-life benefits and risks of active surveillance compared with initial treatment for men with low-risk, clinically localized prostate cancer. DESIGN AND SETTING Decision analysis using a simulation model was performed: men were treated at diagnosis with brachytherapy, intensity-modulated radiation therapy (IMRT), or radical prostatectomy or followed up by active surveillance (a strategy of close monitoring of newly diagnosed patients with serial prostate-specific antigen measurements, digital rectal examinations, and biopsies, with treatment at disease progression or patient choice). Probabilities and utilities were derived from previous studies and literature review. In the base case, the relative risk of prostate cancer-specific death for initial treatment vs active surveillance was assumed to be 0.83. Men incurred short- and long-term adverse effects of treatment. PATIENTS Hypothetical cohorts of 65-year-old men newly diagnosed as having clinically localized, low-risk prostate cancer (prostate-specific antigen level <10 ng/mL, stage ≤T2a disease, and Gleason score ≤6). MAIN OUTCOME MEASURE Quality-adjusted life expectancy (QALE). RESULTS Active surveillance was associated with the greatest QALE (11.07 quality-adjusted life-years [QALYs]), followed by brachytherapy (10.57 QALYs), IMRT (10.51 QALYs), and radical prostatectomy (10.23 QALYs). Active surveillance remained associated with the highest QALE even if the relative risk of prostate cancer-specific death for initial treatment vs active surveillance was as low as 0.6. However, the QALE gains and the optimal strategy were highly dependent on individual preferences for living under active surveillance and for having been treated. CONCLUSIONS Under a wide range of assumptions, for a 65-year-old man, active surveillance is a reasonable approach to low-risk prostate cancer based on QALE compared with initial treatment. However, individual preferences play a central role in the decision whether to treat or to pursue active surveillance.

Critical pathways intervention to reduce length of hospital stay

PURPOSE Despite their popularity, critical pathways have been evaluated in only a few controlled studies. We evaluated the effectiveness of critical pathways in reducing length of hospital stay. SUBJECTS AND METHODS We compared postoperative lengths of stay of patients who underwent coronary artery bypass graft (CABG) surgery, total knee replacement, colectomy, thoracic surgery, or hysterectomy before and after pathway implementation at a university hospital. For three procedures, changes in lengths of stay at neighboring hospitals without pathway programs were assessed for comparison. RESULTS A total of 6,796 patients underwent one of the procedures during the study. The percentage of eligible patients managed on a critical pathway ranged from 94% for hysterectomy to 26% for colectomy. For most procedures, the postoperative length of stay was decreasing during the baseline period. After pathway implementation, the length of stay decreased 21% for total knee replacement, 9% for CABG surgery, 7% for thoracic surgery, 5% for hysterectomy, and 3% for colectomy (all P < 0.01). However, similar decreases were seen in the neighboring hospitals that did not have critical pathways or other specific efficiency initiatives. CONCLUSIONS Critical pathways were associated with a rapid reduction in postoperative length of stay after all five study procedures. Secular trends at nearby hospitals, however, produced comparable reductions for the three procedures available for comparison. These findings raise questions about the effectiveness of critical pathways in a competitive environment.

Triage decisions for emergency department patients with chest pain : do physicians' risk attitudes make the difference ?

OBJECTIVE: To determine whether physicians’ risk attitudes correlate with their triage decisions for emergency department patients with acute chest pain.DESIGN: Cohort.SETTING: The emergency department of a university teaching hospital.PATIENTS: Patients presenting to the emergency department with a chief complaint of acute chest pain.PHYSICIANS: All physicians who were primarily responsible for the emergency department triage of at least one patient with acute chest pain from July 1990 to July 1991.METHODS: The physicians’ risk attitudes were assessed by two methods: 1) a new, six-question risk-taking scale adapted from the Jackson Personality Index (JPI), and 2) the Stress from Uncertainty Scale (SUS).RESULTS: The physicians who had high risk-taking scores (“risk seekers”) admitted only 31% of the patients they evaluated, compared with admission rates of 44% for the medium scorers and 53% for the physicians who had low risk-taking scores (“risk avoiders”), p<0.001. After adjustment for clinical factors, the patients triaged by the risk-seeking physicians had half the odds of admission [odds ratio (OR) 0.51, 95% confidence interval (95% CI) 0.27 to 0.97], and the patients triaged by the risk-avoiding physicians had nearly twice the odds of admission (OR 1.83, 95% CI 1.10 to 3.03) of the patients triaged by the medium-risk scoring physicians. The SUS did not correlate significantly with admission rates. Of the 92 patients released home by the risk-seeking physicians, 91 (99%) were known to be alive four to six weeks afterwards and one was lost to follow-up; among the 66 patients released by the risk-avoiding physicians, 64 (97%) were known to be alive at four to six weeks, one was lost to follow-up, and one died of ischemic heart disease during a subsequent hospitalization (p=NS).CONCLUSIONS: The physicians’ risk attitudes as measured by a brief risk-taking scale correlated significantly with then-rates of admission for emergency department patients with acute chest pain. These data do not suggest that the risk-seeking physicians achieved lower admission rates by releasing more patients who needed to be in the hospital, but an adequate evaluation of the appropriateness of triage decisions of risk-seeking and risk-avoiding physicians will require further study.

The effect of a separate stream for minor injuries on accident and emergency department waiting times

Introduction: To decrease waiting times within accident and emergency (A&E) departments, various initiatives have been suggested including the use of a separate stream of care for minor injuries (“fast track”). This study aimed to assess whether a separate stream of minor injuries care in a UK A&E department decreases the waiting time, without delaying the care of those with more serious injury. Intervention: A doctor saw any ambulant patients with injuries not requiring an examination couch or an urgent intervention. Any patients requiring further treatment were returned to the sub-wait area until a nurse could see them in another cubicle. Method: Data were retrospectively extracted from the routine hospital information systems for all patients attending the A&E department for five weeks before the institution of the separate stream system and for five weeks after. Results: 13 918 new patients were seen during the 10 week study period; 7117 (51.1%) in the first five week period and 6801 (49.9%) in the second five week period when a separate stream was operational. Recorded time to see a doctor ranged from 0–850 minutes. Comparison of the two five week periods demonstrated that the proportion of patients waiting less than 30 and less than 60 minutes both improved (p<0.0001). The relative risk of waiting more than one hour decreased by 32%. The improvements in waiting times were not at the expense of patients with more urgent needs. Conclusions: The introduction of a separate stream for minor injuries can produce an improvement in the number of trauma patients waiting over an hour of about 30%. If this is associated with an increase in consultant presence on the shop floor it may be possible to achieve a 50% improvement. It is recommended that departments use a separate stream for minor injuries to decrease the number of patients enduring long waits in A&E departments.

Importance of Findings on the Initial Evaluation for Cancer in Patients with Symptomatic Idiopathic Deep Venous Thrombosis

The association between advanced cancer and deep venous thrombosis has been well documented, and preexisting visceral cancer is a risk factor for deep venous thrombosis [1, 2]. Some studies [3-7] have suggested that deep venous thrombosis might be a marker for subsequent cancer, but others [8-10] have failed to show such an association. The clinical evaluation for diagnosing underlying cancer in patients with idiopathic deep venous thrombosis remains a challenging problem for physicians [11]. In previous studies of the possible association between deep venous thrombosis and cancer [3, 5, 6, 8], comparison groups comprised patients with secondary deep venous thrombosis (that is, patients with known risk factors [other than cancer] for deep venous thrombosis), patients in whom idiopathic or secondary deep venous thrombosis was suspected but not confirmed [4, 10], or the general population [7, 10]. These studies yielded conflicting results. It would be more appropriate to compare the incidence of cancer in patients who have idiopathic deep venous thrombosis with the incidence of cancer in patients in whom idiopathic deep venous thrombosis had been suspected but was ruled out. Our study was designed to 1) determine the appropriate initial clinical evaluation for cancer in patients with symptomatic idiopathic deep venous thrombosis and 2) estimate the likelihood that cancer will subsequently be diagnosed in patients who had negative results on a clinical evaluation for cancer when deep venous thrombosis occurred. Methods Patient Sample All adults who had venous ultrasonography at Brigham and Womens Hospital (Boston, Massachusetts) between January 1990 and May 1993 because deep venous thrombosis in a lower extremity was suspected were eligible. The study cohort was selected from the daily log of the vascular diagnostic laboratory. Patients were excluded if they presented with a risk factor for deep venous thrombosis [12, 13], such as preexisting cancer; had recently had surgery (within 3 months); had had prolonged immobilization ( 10 days); had recently had trauma to a lower limb (within 3 months); had a documented hypercoagulable state (either deficiency of antithrombin III, protein C, or protein S or the presence of a lupus anticoagulant); were pregnant; had recently given birth (within 2 weeks); or were using oral contraceptives. We also excluded patients who had had a procedure (such as computed tomography of the abdomen) to identify potential underlying cancer before venous ultrasonography, even though the diagnosis of cancer was not known at the time when venous ultrasonography was done. Patients who had had venous ultrasonography as part of the work-up for suspected pulmonary embolism were also excluded because our primary goal was to assess the work-up for cancer in patients with symptomatic idiopathic deep venous thrombosis (such as those with leg pain or swelling), not in patients with chest pain or shortness of breath. If a patient had venous ultrasonography more than once between January 1990 and May 1993, only the first ultrasonogram was considered. The exclusion criteria were chosen before any data were collected. We did not exclude patients with a history of deep venous thrombosis because a recent study [6] showed an increased risk for cancer in patients with recurrent deep venous thrombosis. Of the 5797 venous ultrasonographic examinations, 3445 (59%) were excluded because they had been done in patients who had one or more risk factors (other than cancer) for deep venous thrombosis. Four hundred sixty (8%) of the patients in whom venous ultrasonography was done as part of the work-up for suspected pulmonary embolism were also excluded. Eight hundred thirty patients (14%) were excluded because they had ultrasonography as a follow-up procedure. Fourteen patients (0.2%) were excluded either because a work-up for cancer had been done before venous ultrasonography or because they refused to receive a mailed questionnaire about their medical situation. Finally, 62 patients (1%) were excluded because follow-up information was missing (Figure 1). None of these 62 patients received a diagnosis of deep venous thrombosis. The final cohort comprised 986 patients. The study was approved by the Human Research Committee of Brigham and Womens Hospital. Figure 1. Branching tree of the number of patients included at every stage of the study. Venous Ultrasonography and Definition of Deep Venous Thrombosis In all patients, duplex, color-assisted venous ultrasonography was done in each leg; 5.0 and 7.0 MHz transducers were used as described elsewhere [14]. Diagnosis of deep venous thrombosis was based on the lack of apposition on the venous walls during compression maneuvers. The presence or absence of venous flow augmentation was assessed by color-flow imaging and Doppler waveform analysis (done while the calf muscles were squeezed) to identify the deep venous thrombosis as occlusive or nonocclusive. Findings were confirmed by two observers. Deep venous thrombosis was classified as calf (tibial, peroneal, gastrocnemius, or soleal veins) or proximal (popliteal or femoral veins) venous thrombosis. Venous ultrasonography was used as the gold standard. Data Collection A standardized form was used to collect the following data from the charts of all study patients: date of birth, sex, place of residence, date of procedure, and current or previous use of tobacco. For all patients hospitalized with deep venous thrombosis, a second standardized form was used to collect clinical data from the charts about the initial clinical evaluation. This data included the results of medical history, physical examination, and laboratory tests. This 86-item form was based on the current recommendations for early cancer detection [15, 16] and included 1) findings in the medical history suggestive of underlying cancer, such as weight loss, fatigue, unusual pain, recent modification of bowel or bladder habits, unusual bleeding, or recent modification of cough or hoarseness; 2) findings on the physical examination suggestive of underlying cancer, such as abnormalities on abdominal or breast palpation (abnormal mass), oral cavity examination (ulcerative lesion), skin examination (atypical pigmentation), digital rectal examination (prostatic nodule or irregularities on the rectal walls), or lymphadenopathy or chest auscultation [rales or dullness]; 3) laboratory findings suggestive of underlying cancer, such as positive results on a test for fecal occult blood, abnormal values on the complete blood count and SMA 20, abnormal erythrocyte sedimentation rate, or abnormal results on urinalysis; 4) abnormalities on chest radiography suggestive of underlying cancer, such as lung nodules, pleural effusion, or mediastinal adenopathy; and 5) abnormal results of other procedures, such as computed tomography, ultrasonography, endoscopy, and biopsy. We also recorded the results of laboratory tests that were done after discharge but were related to the work-up done during hospitalization. Knowledge of previous laboratory test results can affect the observation or the recording of clinical findings [17], but we minimized this potential ascertainment bias by collecting data from the medical record only for the first clinical examination. This chart review form also recorded data on alcohol consumption and history of deep venous thrombosis. The person who reviewed the records was blinded to the occurrence of cancer during follow-up. Follow-up and Definition of Cancer The follow-up information for study patients in whom cancer was not diagnosed at the time of referral for ultrasonography was obtained through hospital chart review (38%), computerized record extraction (in patients who were members of a health maintenance organization in the Boston, Massachusetts, area) (54%), mailed questionnaires (6%), telephone interviews (1%), or the death registry of the Massachusetts Registry of Vital Records and Statistics (1%). Follow-up information included occurrence of cancer by site and date (including cancer found at autopsy), date of death or date on which the patient was last known to be alive, cause of death, and occurrence of deep venous thrombosis. Only the first diagnosis of cancer for each patient was considered. All cases of cancer were confirmed by histologic (95%) or cytologic (5%) examination, with the exception of one case of pancreatic cancer for which a clinical diagnosis was accepted. The patient who had this case had a pancreatic mass and liver metastases detected by computed tomography and died 3 weeks after discharge. Cases of nonmelanoma skin cancer were not included. One hundred twenty-two of the 142 patients with deep venous thrombosis (the deep venous thrombosis group) and all of the 844 patients who did not have deep venous thrombosis (the comparison group) who had not received a diagnosis of cancer when deep venous thrombosis was suspected were followed for a median of 34 months (minimum, 3 months [unless the patient died within the first 3 months]; maximum, 60 months). This yielded 304 person-years of follow-up for the deep venous thrombosis group and 2548 person-years of follow-up for the comparison group (Figure 1). Ninety-five percent of the patients in the deep venous thrombosis group and 96% of those in the comparison group were followed for at least 1 year, until death, or until the occurrence of cancer; 2% of the deep venous thrombosis group and 4% of the comparison group were followed for 3 to less than 6 months; and 3% of the deep venous thrombosis group and 2% of the comparison group were followed for 6 to less than 12 months. No follow-up information was obtained for 4 of the 126 patients (3%) in the deep venous thrombosis group. Three of these 4 patients had no findings suggestive of cancer on the initial clinical evaluation when deep venous thrombosis was diagnosed, and the fourth (a 75-year-old man) had only mild anemia, moved to

Cost-effectiveness of PCSK9 Inhibitor Therapy in Patients With Heterozygous Familial Hypercholesterolemia or Atherosclerotic Cardiovascular Disease

IMPORTANCE Proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors were recently approved for lowering low-density lipoprotein cholesterol in heterozygous familial hypercholesterolemia (FH) or atherosclerotic cardiovascular disease (ASCVD) and have potential for broad ASCVD prevention. Their long-term cost-effectiveness and effect on total health care spending are uncertain. OBJECTIVE To estimate the cost-effectiveness of PCSK9 inhibitors and their potential effect on US health care spending. DESIGN, SETTING, AND PARTICIPANTS The Cardiovascular Disease Policy Model, a simulation model of US adults aged 35 to 94 years, was used to evaluate cost-effectiveness of PCSK9 inhibitors or ezetimibe in heterozygous FH or ASCVD. The model incorporated 2015 annual PCSK9 inhibitor costs of

Depression among high utilizers of medical care.

14,350 (based on mean wholesale acquisition costs of evolocumab and alirocumab); adopted a health-system perspective, lifetime horizon; and included probabilistic sensitivity analyses to explore uncertainty. EXPOSURES Statin therapy compared with addition of ezetimibe or PCSK9 inhibitors. MAIN OUTCOMES AND MEASURES Lifetime major adverse cardiovascular events (MACE: cardiovascular death, nonfatal myocardial infarction, or stroke), incremental cost per quality-adjusted life-year (QALY), and total effect on US health care spending over 5 years. RESULTS Adding PCSK9 inhibitors to statins in heterozygous FH was estimated to prevent 316,300 MACE at a cost of

A critical pathway for management of patients with acute chest pain who are at low risk for myocardial ischemia : Recommendations and potential impact

503,000 per QALY gained compared with adding ezetimibe to statins (80% uncertainty interval [UI],