Abstract
The Cleveland Clinic experience with care paths, including their creation and implementation, challenges overcome during development and testing, and outcomes of selected care path evaluations, is described.
Care paths are tools to assist healthcare professionals in practicing evidence-based medicine. The Cleveland Clinic health system has implemented or is developing approximately 100 care paths, including care paths designed to optimize management of sepsis and septic shock and to promote timely use of i.v. tissue plasminogen activator and correct dosing of antithrombotics and statins in patients with stroke. Key steps in successful care path initiatives include (1) identifying key stakeholders, (2) achieving stakeholder consensus on a standardized approach to disease or condition management, (3) cultivating provider awareness of care paths, (4) incorporating care path tools into the electronic health record and workflow processes, and (5) securing the resources to develop, implement, and maintain care paths. Electronic health records facilitate the use of and adherence to care paths. After care path implementation, revisions are typically needed due to unexpected issues not initially identified and to optimize care path features and support resources for clinical practice. Ongoing evaluation is required to determine whether an implemented care path is producing the intended patient and quality performance outcomes.
Care paths provide a standardized approach to treatment or prevention of a disease or condition, reducing unnecessary variability and expense while promoting optimal, cost-effective patient care.
Care paths provide a standardized approach to a disease or condition and reduce unnecessary variability and expense while providing optimal, cost-effective patient care.
Identification of key interprofessional stakeholders, including pharmacists, in care path development is critical for success.
Resource requirements (human and technology) for care paths must be recognized and appreciated, as intensive resource utilization may be the rate-limiting step in creating, implementing, and maintaining care paths at a hospital or within a health system.
Healthcare is moving from a volume-based reimbursement model to a value-based model (pay for performance) whereby value is defined as outcomes divided by cost.1 Payment is shifting from fee-for-service to outcomes-based payment, the focus of healthcare is moving from acute episode management to population health management, and the role of providers is changing from managing patients at single encounters to working in teams to manage patients across a care continuum (a bundled approach to care). To manage the paradigm shift, clinicians need tools to guide evidence-based practice while providing quality care with minimal cost. These tools include protocols, guidelines, and care paths. Care paths are tools to assist healthcare professionals in practicing evidence-based medicine. Care paths should provide an interprofessional care approach, contain evidenced-based interventions and plans, outline specifics of the care plan, establish timelines, and standardize overall care for a specific population, condition, or disease.2,–4 More specifically, care paths should have explicit, prioritized interventions incorporated into a clinical workflow vetted by an interprofessional team.5 Ideally, care paths should rely not on clinicians but on systems (e.g., an electronic health record) to facilitate care path use, with the caveat of information technology being resource intensive.5 Pharmacists have a unique perspective and expert opinion on the use of medications in care paths. They can identify medication-related issues, either clinically or operationally, not recognized by physicians, nurses, or other healthcare professionals. Therefore, active participation by pharmacists, as part of the interprofessional team, in the construction, implementation, and evaluation of care paths is imperative. This article describes the use of care paths in a large health system.
Cleveland Clinic comprises a 1,400-bed academic medical center, 9 community hospitals in northeast Ohio, and Cleveland Clinic Florida, with a total of 4,450 beds in the health system. Approximately 3,500 physicians and scientists and over 14,000 nurses are employed by Cleveland Clinic. The pharmacy department has over 1,000 full-time employees, including pharmacists, technicians, and support personnel. Cleveland Clinic has a single electronic health record across the enterprise for inpatients and outpatients. Since 2010, Cleveland Clinic has had a health-system pharmacy and therapeutics committee and integrated formulary. In 2016, Cleveland Clinic had 158,200 admissions and 4.2 million outpatient clinic visits, the Case Mix Index was 1.88, and the health-system drug budget, which encompasses inpatient, outpatient clinic, ambulatory care, and specialty pharmacies, exceeded $775 million.
Care paths have been used at Cleveland Clinic since 2010. The care paths can be disease or condition based (e.g., knee replacement, low back pain) or process based (e.g., venous thromboembolism prevention, preadmission testing). There are 3 stages to Cleveland Clinic’s care path development and implementation: (1) the “guide” stage, (2) the pilot stage, and (3) the “build” (deployment) stage. The guide stage involves creating an interprofessional team, reviewing published literature and guidelines, preparing a written document containing quality- and cost-driven care recommendations, and securing caregiver approval. In the pilot stage, operational pilot tests of care path documents are conducted on a small scale, with limited development and resource utilization, allowing for care path evaluation and assessment. A care path build is a large-scale deployment with full integration into the electronic health record and includes performance metrics to assess value; this is the most resource-intensive stage.
Cleveland Clinic currently has 100 care path guides available for health-care professionals on an internal website. Of these care path guides, 40% are in pilot status and 20% are in build status. Cleveland Clinic uses an online dashboard to track the status of care path development and implementation (Table 1). Two fully completed and implemented care paths focus on stroke care and treatment of sepsis and septic shock. This article reviews the specific care process established for each of these care paths and provides insights on their creation and implementation, obstacles encountered and solutions devised, and outcomes of care path–guided treatment.
Status of Selected Care Paths in Use or in Development by Cleveland Clinic Health System
| Care Path Focus | Guide Development | Pilot Testing | Deployment |
|---|---|---|---|
| Abdominal surgery (enhanced recovery after surgery) | In progress | Completed | In progress |
| Adult migraine | Completed | Completed | Activity pendinga |
| Cervical cancer | Completed | Activity pending | Activity pending |
| Chronic obstructive pulmonary disease (acute exacerbation) | Completed | In progress | In progress |
| Colon cancer | Completed | Completed | Completed |
| Hospital-acquired pneumonia | Completed | Completed | Activity pending |
| Multiple myeloma | Completed | Completed | No activity |
| Prostate cancer | Completed | Completed | In progress |
| Severe sepsis | Completed | Completed | Completed |
| Stroke | Completed | Completed | Completed |
| Total hip arthroplasty | Completed | Completed | Completed |
| Total knee arthroplasty | Completed | Completed | Completed |
| Care Path Focus | Guide Development | Pilot Testing | Deployment |
|---|---|---|---|
| Abdominal surgery (enhanced recovery after surgery) | In progress | Completed | In progress |
| Adult migraine | Completed | Completed | Activity pendinga |
| Cervical cancer | Completed | Activity pending | Activity pending |
| Chronic obstructive pulmonary disease (acute exacerbation) | Completed | In progress | In progress |
| Colon cancer | Completed | Completed | Completed |
| Hospital-acquired pneumonia | Completed | Completed | Activity pending |
| Multiple myeloma | Completed | Completed | No activity |
| Prostate cancer | Completed | Completed | In progress |
| Severe sepsis | Completed | Completed | Completed |
| Stroke | Completed | Completed | Completed |
| Total hip arthroplasty | Completed | Completed | Completed |
| Total knee arthroplasty | Completed | Completed | Completed |
“Activity pending” denotes that work had not been planned or started at the time of writing.
Status of Selected Care Paths in Use or in Development by Cleveland Clinic Health System
| Care Path Focus | Guide Development | Pilot Testing | Deployment |
|---|---|---|---|
| Abdominal surgery (enhanced recovery after surgery) | In progress | Completed | In progress |
| Adult migraine | Completed | Completed | Activity pendinga |
| Cervical cancer | Completed | Activity pending | Activity pending |
| Chronic obstructive pulmonary disease (acute exacerbation) | Completed | In progress | In progress |
| Colon cancer | Completed | Completed | Completed |
| Hospital-acquired pneumonia | Completed | Completed | Activity pending |
| Multiple myeloma | Completed | Completed | No activity |
| Prostate cancer | Completed | Completed | In progress |
| Severe sepsis | Completed | Completed | Completed |
| Stroke | Completed | Completed | Completed |
| Total hip arthroplasty | Completed | Completed | Completed |
| Total knee arthroplasty | Completed | Completed | Completed |
| Care Path Focus | Guide Development | Pilot Testing | Deployment |
|---|---|---|---|
| Abdominal surgery (enhanced recovery after surgery) | In progress | Completed | In progress |
| Adult migraine | Completed | Completed | Activity pendinga |
| Cervical cancer | Completed | Activity pending | Activity pending |
| Chronic obstructive pulmonary disease (acute exacerbation) | Completed | In progress | In progress |
| Colon cancer | Completed | Completed | Completed |
| Hospital-acquired pneumonia | Completed | Completed | Activity pending |
| Multiple myeloma | Completed | Completed | No activity |
| Prostate cancer | Completed | Completed | In progress |
| Severe sepsis | Completed | Completed | Completed |
| Stroke | Completed | Completed | Completed |
| Total hip arthroplasty | Completed | Completed | Completed |
| Total knee arthroplasty | Completed | Completed | Completed |
“Activity pending” denotes that work had not been planned or started at the time of writing.
Stroke care path
Cleveland Clinic is a Joint Commission–certified comprehensive stroke center with more than 3,200 visits annually, 1 of the highest patient volumes in North America.6 There are established guidelines to direct initial management of ischemic stroke as well as multiple requirements for Joint Commission certification.7 Therefore, stroke care was a natural choice for implementation of the first care path at Cleveland Clinic. The goals of the stroke care path, implemented in September 2010, are to (1) improve efficiency and quality in management of stroke and (2) collect information in structured fields of the electronic health record (e.g., stroke quality metrics, patient-specific variables, patient-reported outcomes) to allow use of clinical data in aggregate analyses for research, operational, and quality-improvement activities.
One of the first steps in the stroke care path development process was obtaining consensus among providers on the basic tenets of managing stroke to build algorithms related to stroke care. This critical initial step fostered an environment for provider acceptance of the care path to ultimately optimize its utilization. This was initially a challenge, as providers were concerned about the potential loss of autonomy in patient management. To alleviate this apprehension, it was reinforced that the care path was created as a guide to improve efficiency and help providers adhere to the quality indicators relevant to the majority of patients with stroke, allowing providers to focus more attention on the multitude of items requiring clinical judgment. After reaching consensus on the care path guide, algorithms were developed to provide direction on each step in care and then the algorithms were incorporated in the electronic health record in the form of order sets, best-practice alerts, and documentation tools. The use of order sets not only guides providers in evidence-based stroke management but decreases the time required for placing orders, because commonly used orders for patient care or necessary orders for quality metrics are in a central location: the order set. For example, dysphagia screening, which is a quality measure and essential to determine the appropriate route of medication administration, is preselected in order sets to ensure adherence. Additionally, these order sets assist providers in the correct dosing of i.v. tissue plasminogen activator (Figure 1) and the ordering of medications such as antithrombotic and statin therapies. Providers are also able to place an order for a postdischarge follow-up appointment, significantly reducing the time it takes to manually set up an outpatient appointment, which previously required 1 or more phone calls. The documentation tools incorporated in the care path improve the efficiency of collecting essential data for stroke care, such as National Institutes of Health Stroke Scale scores and metrics required by the Centers for Medicare and Medicaid Services (CMS), by providing context-driven questions, which reduces unnecessary data entry, and by automatically populating stroke-specific clinical notes in the electronic health record. The documentation tools also serve as clinical aids. For instance, a documentation checklist for i.v. tissue plasminogen activator reminds providers of eligibility criteria for use in acute ischemic stroke. A stroke care checklist that is prepopulated with information from the electronic health record provides the status of several clinical parameters that should be monitored daily in patients admitted for ischemic stroke, such as nutrition orders and blood pressure.8 Providers can document their current treatment strategy as they rapidly assess the current status of these items. Because much of the information in the stroke care path is collected in a structured format, the data can be extracted at a later date for multiple uses such as measuring adherence to key metrics required for Joint Commission certification. In addition, patient-reported outcomes are collected at the time of postdischarge follow-up visits through electronic questionnaires, which provide information on important outcomes after stroke such as physical function, depressive symptoms, fatigue, and satisfaction with social roles. These types of data are used for both clinical management and research purposes.
Screenshot of electronic health record (Epic Systems Corporation, Verona, WI) of criteria for use of i.v. tissue plasminogen activator as part of the Cleveland Clinic stroke care path.
A formal analysis of the impact of the stroke care path demonstrated a significant reduction in the ratio of observed to expected inpatient mortality and length of stay in patients with ischemic stroke in the first 2 years after implementation, as compared with the period immediately prior to implementation. This suggested that the care path may be associated with improved outcomes in patients with ischemic stroke.9 There was no difference in adherence to performance measures during the preimplementation and postimplementation periods, although the rate of adherence was above 95% in both periods, limiting the ability to detect improvement.9
Several challenges were encountered throughout the creation and implementation of the stroke care path. First, there was a lack of use or incorrect use of the care path if providers, especially rotating residents, were unfamiliar with the tool. As a solution, regular provider and trainee education is a continuing initiative to optimize the appropriate utilization of the stroke care path. Second, there were functionality limitations within the electronic health record, and some planned features such as an autopopulated discharge checklist did not work as anticipated. Providing users with the ability to easily extract structured stroke care path data from the electronic health record continues to be a challenge.
In 2014 the stroke care path underwent a revision process, which included updates in the treatment of stroke and the addition of the management of intracerebral hemorrhagic and subarachnoid hemorrhage. The updated care path also improved processes for extracting data to generate a dashboard for quality measures and enabled on-demand reporting to identify patients lost to follow-up. Use of the stroke care path has been expanded to all the hospitals in the Cleveland Clinic health system.
Sepsis and septic shock care path
The sepsis and septic shock care path was developed by Cleveland Clinic’s Respiratory Institute, which provides critical care services to the 64-bed medical intensive care unit (ICU) where patients with these conditions are primarily treated. Cleveland Clinic treats over 3,500 patients with sepsis per year, with an in-hospital mortality rate above 20%. A care path was designed for this disease state in order to standardize care in the acute phases, with less emphasis placed on treatment across continuums of care. In 2013, a sepsis and septic shock care path guide drafting committee was formed; it consisted of the care path sponsor (a physician leader from the Respiratory Institute), an additional Respiratory Institute physician practicing primarily in the ICU, a Respiratory Institute nurse, a quality-improvement nurse, an ICU pharmacist, and a process improvement specialist. The resulting care path is a specific example of how care paths can be used for the management of limited time frame events (e.g., surgical procedures, pneumonia) as opposed to conditions requiring longitudinal care (e.g., asthma, chronic obstructive pulmonary disease, stroke). The sepsis and septic shock care path incorporates patient identification tools and workflow optimization strategies (e.g., autopopulated laboratory test orders) into a comprehensive treatment plan. Development of the care path guide was based on national or international guidelines.10,11 Importantly, process maps were included in the care path guide to outline the optimal workflow for patient screening and identification and implementation of treatment bundles (Figures 2 and 3). After the initial draft was sent out for review, it was clear that a close partnership with the Emergency Services Institute (which provides service to emergency departments across the Cleveland Clinic health system) was vital to the success of the care path. Frequent meetings were held to outline best practices for patient identification, managing throughput from the emergency department to the ICU or ward, and to align treatment practices. After 2 rounds of review, the care path guide was approved, the order set was scheduled for building, and a pilot test was designed.
Bundle intervention process specified in Cleveland Clinic sepsis and septic shock care path. ICU = intensive care unit, RNF = regular nursing floor, ED = emergency department.
Screening process specified in Cleveland Clinic sepsis and septic shock care path. ICU = intensive care unit, UA = urinalysis, CXR = chest x-ray, WBC = white blood cell, SBP = systolic blood pressure, MAP = mean arterial pressure, PaO2 = partial pressure of oxygen, FiO2 = fraction of inspired oxygen, Cr = creatinine, INR = International Normalized Ratio, PTT = partial thromboplastin time, POC = point of care, CBC = complete blood count, ABG = arterial blood gas, CT = computed tomography, ED = emergency department, RNF = regular nursing floor.
The care path order set was prioritized as one of the first order sets built and applied in light of the high case mortality rates associated with sepsis and septic shock. During the order set design, an important challenge was encountered. Because the Cleveland Clinic system uses a single electronic health record, alignment of shared orders (including all medication records) was required for the order set. A key antibiotic in the care path is piperacillin–tazobactam, which at the time of care path development was administered by intermittent i.v. bolus injection or extended infusion in different hospitals. A task force was charged with recommending the optimal administration technique; in light of the data available at the time, intermittent bolus administration was recommended. This recommendation was accepted by the chief pharmacy officer and individual hospital pharmacy directors and implemented for all piperacillin–tazobactam orders across the health system. Importantly, despite the merit of both administration approaches, consensus was gained and processes were aligned in order to facilitate implementation of the care path.
As the order set was being built and enhanced, a pilot test of the care path was planned and conducted. In the pilot design phase, a challenge identified involved serum lactate ordering, including the processes for sending blood samples to the laboratory for blood gas versus chemistry analysis and issues of laboratory processing priority and reporting method. Notably, with the assistance of the care path steering committee, each Cleveland Clinic hospital developed individualized processes to ensure the most prompt reporting of lactate values. In early 2014, the pilot test was conducted at the academic medical center and 1 community hospital. This pilot test was primarily initiated in the emergency departments and in 2 ward units at these institutions, with continued care, if necessary, provided in the ICU setting. Many patient care metrics improved during care path pilot testing; most notably, emergency department length of stay was decreased by an average of 78 minutes, 11% more patients were eventually discharged home instead of to a skilled nursing facility, and the average cost per case was decreased by 17%. Based on these favorable findings, the sepsis and septic shock care path was sequentially implemented across the Cleveland Clinic health system in late 2014 and 2015.
The care path has evolved over time, particularly because of the implementation of a CMS quality measure for severe sepsis and septic shock care (core measure SEP-1) in 2016. Also, new definitions of sepsis, with terminology changes, have generated the need to provide continued education and feedback to all caregivers. Cleveland Clinic’s Enterprise Sepsis Steering Committee continues to meet monthly to discuss metrics, optimize workflow, and address concerns.
Discussion
The Cleveland Clinic experience has imparted lessons from which recommendations for successful care path initiatives can be made. We have found a 3-stage approach (guide development, pilot testing, and deployment) to be optimal. Assessment of the care path is an additional step that should be performed after care path implementation.
In the first stage, during which the basic structure of a care path is created, consider all potential stakeholders for review of the care path and obtain consensus on the evidence-based management of the disease or condition. The interprofessional stakeholders should review published literature and guidelines, gain consensus on a standardized approach to patient care, and prepare a written document containing quality- and cost-driven care recommendations; this was one lesson learned in implementing the sepsis and septic shock care path. Few infectious diseases practitioners were formally included in the review process, and this led to criticism of how some guideline recommendations were operationalized.
During the second stage, the care path should be pilot tested on a small scale, with limited development and resource utilization. Areas for improvement or modification can be identified prior to care path implementation on a larger scale. Pilot testing can detect a number of operational opportunities, such as misalignment of electronic workflows and proposed workflows. During testing of the stroke care path, some anticipated functionalities, such as the discharge checklist, did not work in clinical practice. As another example, the original version of the sepsis and septic shock care path order set specified a 1-time order for broad-spectrum antibiotic. This design element was chosen in order to allow providers to utilize broad-spectrum antimicrobials initially but select the most appropriate antibiotic therapy (possibly one with a narrower spectrum of activity) for subsequent orders. However, despite intensive education, there were several reports of patients inadvertently missing antibiotic doses because providers erroneously thought prescriptions for broad-spectrum agents made through the order set were handled as ongoing scheduled orders. In light of these findings, the default 1-time frequency for these antibiotic orders was removed from the order set, and providers were required to specify the desired frequency.
During the care path deployment stage, provider education and awareness are crucial for success. If providers do not know the care path exists or do not use the care path appropriately, the intended positive impact on patient care may not be seen. In addition, patient care may not be optimal, and anticipated efficiencies in provision of care may not be realized. Furthermore, providers need to be educated about care paths in general. Broad education on care paths should address topics such as (1) definitions and purpose, (2) the creation process and stakeholder involvement, (3) strategies for addressing provider concerns of loss of autonomy and diminished reliance on clinical judgment in patient care, and (4) measurements of care path success, including patient outcomes and quality metrics. Stroke care path education was conducted by the Cleveland Clinic cerebrovascular center stroke team. As stated earlier in the article, education should be done on an as-needed basis, not only at the time of care path implementation. Sepsis and septic shock care path education occurred at each hospital and was led by individuals in quality improvement and nursing education. There was a particular emphasis on educating in hospital areas where sepsis was more prevalent, such as medicine wards and ICUs.
After the build stage, team members should determine if the care path has yielded the intended results. Important questions include the following: Were patient outcomes improved? Was efficiency realized, or were unintended inefficiencies introduced in the system? Did the care path enhance and capture relevant quality metrics? Based on the answers to these questions and other feedback, the care path may need to be revised. At Cleveland Clinic, quality metrics related to care paths are integrated into quality dashboards and are reviewed at least annually by the institutional leadership. At Cleveland Clinic, there are no specific estimates of cost savings associated with use of care paths; however, the anticipated return on investment for most care paths is decreased length of stay, readmissions, and secondary events, leading to overall cost reduction.
Resource requirements (human and technological) for care paths must be recognized and appreciated. Intensive resource utilization may be the rate-limiting step in creating, implementing, and maintaining care paths at a hospital or within a health system. Potential solutions might include creative use of available resources and allowance of dedicated time to participate in care path–related activities. Additionally, if the informatics pharmacists report directly to the department of pharmacy (and not the hospital’s information technology department), direction of more time toward care path development and maintenance may be feasible.
Finally, pharmacists can play an integral role in care paths. Pharmacists need to be incorporated as part of the interprofessional team when a care path is being created. They can assist in ensuring that cost-effective medications are included. Additionally, pharmacists can provide input for medication-related issues (clinical or operational), specifically when care path content is translated to order sets. They can help provide education regarding the availability of care paths and corresponding order sets, appropriate care paths to use, and adherence to the care paths. Finally, pharmacists can review data and outcomes collected and recommend appropriate revisions to care paths.
Conclusion
Care paths provide a standardized approach to treatment or prevention of a disease or condition, reducing unnecessary variability and expense while promoting optimal, cost-effective patient care.
Disclosures
The authors have declared no potential conflicts of interest.
Acknowledgments
The authors thank and acknowledge Guido Bergomi, B.S., for assistance in obtaining select Cleveland Clinic care path data.
Footnotes
This article is part of a special AJHP theme issue on population health management. Contributions to this issue were coordinated by AJHP Editorial Advisory Board member Kenneth M. Shermock, Pharm.D., Ph.D.
References
Comments