BHA 320 MGT of Health Programs

Mod4readCh4.pdf

Home >Nursing homework help >BHA 320 MGT of Health Programs

T h e H e a l t h c a r e Q u a l i t y B o o k92

All Baldrige applicants receive a feedback report evaluating the strengths and weaknesses of their responses to each of the seven categories. The purpose of the feedback report is to document the analysis of the appli- cant’s response so that it can be used to evaluate the organization’s responses to future applications and identify potential gaps in the organization’s strate- gic planning and improvement activities.

The national Baldrige criteria serve as the framework for many state and local quality awards. In 2012, eligibility requirements for the Baldrige Award were changed; applicants now must have received a “top-tier award” from a state or local Baldrige-based award program or meet one of five condi- tions related to past national or state-based award performance.

Lean/Toyota Production System The Massachusetts Institute of Technology developed the term Lean in 1987 to describe product development and production methods that, when com- pared with traditional mass production processes, produce more products with fewer defects in a shorter time. The goal was to develop a way to specify value, align steps/processes in the best sequence, conduct these activities with- out interruption whenever someone requests them, and perform them more effectively (Womack and Jones 2003). Lean thinking, sometimes called Lean manufacturing or the Toyota Production System (TPS), focuses on the removal of waste (muda), which is defined as anything that is not needed to produce a product or service. Taiichi Ohno (cofounder of TPS) identified seven types of waste: (1) overproduction, (2) waiting, (3) unnecessary transport, (4) over- processing, (5) excess inventory, (6) unnecessary movement, and (7) defects.

The focus of Lean methodology is a “back to basics” approach that places the needs of the customer first through the following five steps:

1. Define value as determined by the customer, identified by the provider’s ability to deliver the right product or service at an appropriate price.

2. Identify the value stream, the set of specific actions required to bring a specific product or service from concept to completion.

3. Make value-added steps flow from beginning to end. 4. Let the customer pull the product from the supplier; do not push

products. 5. Pursue perfection of the process.

Although Lean focuses on removing waste and improving flow, it also has some secondary effects. Quality is improved. The product spends less time in process, reducing the chances of damage or obsolescence. The simplification of processes reduces variation and inventory and increases the uniformity of outputs (Heim 1999).

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

EBSCO Publishing : eBook Academic Collection (EBSCOhost) - printed on 2/28/2018 3:04 PM via TRIDENT UNIVERSITY AN: 863699 ; Joshi, Maulik, Ransom, Elizabeth R., Nash, David B., Ransom, Scott B..; The Healthcare Quality Book : Vision, Strategy, and Tools Account: s3642728

93C h a p t e r 4 : Q u a l i t y I m p r o v e m e n t 93

Six Sigma Six Sigma (3.4 defects per million) is a system for improvement developed by Hewlett-Packard, Motorola, General Electric, and others over the course of the 1980s and 1990s (Pande, Neuman, and Cavanagh 2000). The tools used in Six Sigma are not new. The thinking behind this system builds on the foundations of quality improvement established in the 1930s through the 1950s. What makes Six Sigma appear new is the rigor of tying improve- ment projects to key business processes and clear roles and responsibilities for executives, champions, master black belts, black belts, and green belts.

The aim of Six Sigma is to reduce variation (eliminate defects) in key business processes. By using a set of statistical tools to understand the fluctuation of a process, managers can predict the expected outcome of that process. If the outcome is not satisfactory, management can use associated tools to learn more about the elements influencing the process. Six Sigma includes five steps—define, measure, analyze, improve, and control—com- monly known as DMAIC:

1. Define: Identify the customers and their problems. Determine the key characteristics important to the customer along with the processes that support those key characteristics. Identify existing output conditions along with process elements.

2. Measure: Categorize key characteristics, verify measurement systems, and collect data.

3. Analyze: Convert raw data into information that provides insights into the process. These insights include identifying the fundamental and most important causes of the defects or problems.

4. Improve: Develop solutions to the problem, and make changes to the process. Measure process changes, and judge whether the changes are beneficial or another set of changes is necessary.

5. Control: If the process is performing at a desired and predictable level, monitor the process to ensure that no unexpected changes occur.

The primary theory of Six Sigma is that a focus on reducing variation leads to more uniform process output. Secondary effects include less waste, less throughput time, and less inventory (Heim 1999).

Quality tools

One of the difficult things about quality is explaining how a tool is different from a process or a system. We can observe people using tools and methods for improvement. We can see them making a flowchart, plotting a control chart, or using a checklist. These tools and procedures are the logical outcomes

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k94

of process and system changes that people have put in place or implemented to make improvements or identify a problem. People may use several tools and procedures to make improvements, and these tools may form one part of an improvement system. Although we can observe people using the tools of the system, the system (e.g., Six Sigma, Lean) itself is invisible and cannot be observed. Many of the more than 50 quality tools available today were devel- oped to “see” the quality system they are designed to support. The American Society for Quality (Tague 2004) has classified quality tools into six categories:

1. Cause analysis 2. Evaluation and decision making 3. Process analysis 4. Data collection and analysis 5. Idea creation 6. Project planning and implementation

This section of the chapter is not intended to be a comprehensive reference on quality tools and techniques but rather highlights some of the more widely used tools. The following discussion organizes the tools into three categories:

1. Basic quality tools 2. Management and planning tools 3. Other quality tools

Basic Quality Tools Basic quality tools are used to define and analyze discrete processes that usually produce quantitative data. These tools primarily are used to explain a process, identify potential causes for process performance problems, and collect and display data indicating which causes are most prevalent.

5 whys Simple to understand and perform, the 5 Whys exercise was developed as a basic method for drilling down through the symptoms of a process or design failure to identify the root cause. By asking why or what caused the problem, users of this technique can quickly identify possible root causes and make improvements that will correct the real problem, not just address the symp- toms. Key to successful use of this technique is not to stop the analysis too early so as to misidentify the root cause.

Control Chart Also referred to as statistical process control, control charts are graphs used to display data for the purpose of identifying how processes or outcomes change

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

95C h a p t e r 4 : Q u a l i t y I m p r o v e m e n t 95

over time. Control charts contain three lines: a central/control line (aver- age), an upper control limit, and a lower control limit. These boundaries are used to measure and monitor performance to identify performance tenden- cies and variation. Control charts also can be used to assess the impact of a process change on performance, enabling the user to correct or identify any problems that arise (Tague 2004).

Histogram A histogram is a graphical display of the frequency distribution of a quality characteristic of interest. A histogram makes variation in a group of data apparent and aids analysis of the distribution of data around an average or median value.

Cause-and-effect/fishbone Diagram Cause-and-effect diagrams are sometimes referred to as Ishikawa, or fish- bone, diagrams. In a cause-and-effect diagram, the problem (effect) is stated in a box on the right side of the chart, and likely causes are listed around major headings (bones) that lead to the effect. Cause-and-effect diagrams can help organize the causes contributing to a complex problem (ASQ 2014).

Pareto Chart Vilfredo Pareto, an Italian economist in the 1880s, observed that 80 percent of the wealth in Italy was held by 20 percent of the population. Juran later applied this principle to other applications and found that 80 percent of the variation of any characteristic is caused by only 20 percent of the possible variables. A Pareto chart is a display of occurrence frequency that shows this small number of significant contributors to a problem, enabling management to concentrate resources and identify the frequency with which specific errors are occurring (Tague 2004).

Checksheet Checksheets are a generic tool designed for multiple data-collection purposes. They are used to capture data measured repeatedly over time for purposes of identifying patterns, trends, defects, or causes of defects. Data collected using a checksheet can be easily converted into data performance tools such as histograms or Pareto charts (Tague 2004).

Management and Planning Tools Managers use management and planning tools to organize the decision- making process and create a hierarchy when faced with competing priorities. These tools also are useful for dealing with issues involving multiple depart- ments in an organization and for creating an organization-wide quality culture.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k96

Balanced scorecard Renowned management consultant Peter Drucker is often quoted as having said “you can’t manage what you don’t measure.” Developed by Dr. Robert Kaplan and Dr. David Norton, the balanced scorecard is used to collect, measure, and analyze the strategic planning and management of an organization. This tool transfers high-level organizational performance expectations to the individual department level to measure the impact of day-to-day operations and deliver- ables. Through visual display of performance measures in the areas of finance, customers, internal (business) processes, and employee learning and growth, an organization can reinforce its priorities and design specific systems and processes around its vision and strategy (Balanced Scorecard Institute 2014).

affinity Diagram Affinity diagrams can encourage people to develop creative solutions to problems. For example, the use of an affinity diagram is a way to create order out of a brainstorming session. An issue or problem is identified, and then individuals record their own ideas about the issue/problem on small note cards. As a group, team members study the cards and then group the recorded ideas into common categories.

Matrix relations Diagram The matrix relations diagram helps us answer two important questions when sets of data are compared: (1) Are the data related? and (2) How strong is the relationship? The House of Quality, a quality function deployment tool, is an example of a matrix relations diagram. It lists customers’ needs on one axis and an organization’s/product’s capabilities on the second axis. The diagram compares what the customer wants with how the vendor will meet those expectations. The matrix relations diagram can identify not only relation- ships between sets of data but also patterns in the relationships and serves as a useful checklist for ensuring that tasks are being completed (Tague 2004).

stratification When gathering data from multiple sources or conditions, researchers may use the technique of stratification to analyze and determine whether data variation exists among the sources. Stratification can help researchers identify patterns in the data and prevent misrepresentation of study findings when data from multiple sources are presented together.

scatter Diagram Scatter diagrams enable users to identify whether a correlation exists between pairs of numerical data. Also known as a scatter plot or X-Y graph, the scatter diagram can be used in a root cause analysis to determine the cause-and-effect

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

97C h a p t e r 4 : Q u a l i t y I m p r o v e m e n t 97

relationship that two elements may have. The greater the correlation between the two elements, the more the data will display as a tight line or curve, whereas two disparate elements will display as a more scattered or “shotgun” distribution.

Priorities Matrix Use of a priorities matrix involves the application of a series of planning tools built around the matrix chart. When tasks outnumber available resources, managers can use this matrix to prioritize work on the basis of data rather than emotion. Priorities matrixes enable managers to systematically discuss, identify, and prioritize the criteria that most influence their decisions about which tasks to complete and to study different possibilities for prioritizing tasks (ASQ 2014).

Other Quality Tools Benchmarking Organizations use benchmarking to compare the processes and successes of their competitors or of similar top-performing organizations to their own processes to identify process variation and organizational opportunities for improvement.

failure Mode and effects analysis Failure mode and effects analysis (FMEA) examines potential problems and their causes and predicts undesired results. FMEA normally is used to pre- dict product failure from past part failure, but it also can be used to analyze future system failures. This method of failure analysis generally is performed on product design and work processes. By basing their activities on FMEA, organizations can focus their efforts on steps in a process that have the great- est potential for failure before failure actually occurs. Prioritization of failure modes to address and mitigate is based on the detectability of the potential failure, its severity, and its likelihood of occurrence.

flowchart Flowcharts are used to visually display the steps of a process in sequential order. Each step in a flowchart is displayed as a symbol that represents a particular action (e.g., process step, direction, decision, delay). For quality improvement purposes, flowcharts are useful tools for identifying unneces- sary steps in a process, developing procedures, and facilitating communica- tion between staff involved in the same process (Tague 2004).

spaghetti Diagram First developed in the manufacturing industry to display the path of an item through a factory, spaghetti diagrams are used to identify unnecessary

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k98

repetition in a process and opportunities for improved efficiency (i.e., removal of unnecessary steps). By visually displaying multiple simultaneous processes, spaghetti diagrams can reveal potential causes of delay or unneces- sary motion.

5s The Japanese tool 5S (each step starts with the letter “S”) is a systematic program that helps workers take control of their workspace so that it helps them complete their jobs instead of being a neutral or, as is commonly the case, a competing factor:

1. Seiri (sort) means to keep only items necessary for completing one’s work.

2. Seiton (straighten) means to arrange and identify items so that they can be easily retrieved when needed.

3. Seiso (shine) means to keep items and workspaces clean and in working order.

4. Seiketsu (standardize) means to use best practices consistently. 5. Shitsuke (sustain) means to maintain gains and make a commitment to

continue to apply the first four Ss.

Mistake Proofing (Poka yoke) A concept developed in the 1960s by Japanese industrial engineer and TPS cofounder Shigeo Shingo, mistake proofing is the creation of techniques and devices to ensure that processes work right from the first time they are implemented. Mistake proofing techniques can be used to address potential failures identified during FMEA. The goal of mistake proofing is to make an error impossible to occur or easily detectable before significant consequences result.

Knowledge transfer and spread techniques

A key aspect of any quality improvement effort is the ability to replicate suc- cesses in other areas of the organization. Barriers to spread and adoption (e.g., organizational culture, communication, leadership support) exist in any unit, organization, or system. However, failure to transfer knowledge effectively may cause an organization to produce waste, perform inconsis- tently, and miss opportunities to achieve benchmark levels of operational performance.

The concept of transfer of learning, developed in 1901, explores how individuals can apply lessons learned in one context to another context. The

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

99C h a p t e r 4 : Q u a l i t y I m p r o v e m e n t 99

theory relies on the notion that the characteristics of the new setting are similar enough to those of the previous setting that processes can be repli- cated and similar efficiencies can be gained in the new setting (Thorndike and Woodworth 1901).

In 1999, the Institute for Healthcare Improvement (IHI) chartered a team to create a “framework for spread.” In 2006, IHI published “A Framework for Spread: From Local Improvements to System-Wide Change,” a white paper that identified “the ability of healthcare providers and their organizations to rapidly spread innovations and new ideas” as a “key factor in closing the gap between best practice and common practice” (Massoud et al. 2006, 1). The report noted the following questions as important for orga- nizations to address when attempting to spread ideas to their target popula- tions (Massoud et al. 2006, 6):

• Can the organization or community structure be used to facilitate spread?

• How are decisions about the adoption of improvements made?

• What infrastructure enhancements will assist in achieving the spread aim?

• What transition issues need to be addressed?

• How will the spread efforts be transitioned to operational responsibilities?

The following discussion presents techniques that can be used to facilitate spread within a department, across an organization, or throughout a system. The decision to use any of these techniques depends on the goals and complexity of the changes to be disseminated. Like the group of quality improvement systems and tools presented earlier in the chapter, this selection of knowledge transfer techniques is only a representative sample of the many methods available for this purpose.

Kaizen Blitz/Event Kaizen, translated as “continuous improvement,” was developed in Japan shortly after World War II. Kaizen in any organization involves ongoing improvement that is supported and implemented at all levels of an organiza- tion. The key aspect of Kaizen is the continual focus on improving a system or process regardless of how well the system or process is currently function- ing. A Kaizen “blitz” or event is a highly focused improvement effort aimed at addressing a specific problem. Kaizen events are short in duration—typi- cally three to five days. As such, Kaizen blitzes/events are intended to pro- duce rapid changes that produce quick results. The approach to improvement taken during a Kaizen blitz/event typically involves common improvement methodologies (e.g., DMAIC, PDCA, value stream mapping) and the partic- ipation of teams with decision-making authority from multiple departments and levels of leadership.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k100

Rapid-Cycle Testing/Improvement Two important characteristics of an effective spread model are staff buy-in and proof that the change will improve performance. Developed by IHI, rapid-cycle testing (or rapid-cycle improvement) was designed to create vari- ous small tests involving small sample sizes and multiple PDSA cycles that build on the lessons learned in a short period while gaining buy-in from staff involved in the change (see Exhibit 4.3). Successful tests are applied to other units in the organization, whereas unsuccessful tests continue to be revised for potential spread and further implementation. Rapid-cycle testing is designed to reduce the cycle time of new process implementation from months to days. To prevent unnecessary delays in testing or implementation, teams or units using rapid-cycle testing must remain focused on testing solu- tions and avoid overanalysis. Rapid-cycle testing can be resource intensive (i.e., involves high resource consumption in a short period) and therefore may require top-level leadership support.

Case Study: Reengineering Discharge in a Community-Wide Collaborative Project to Reduce Hospital Readmissions In August 2008, TMF Health Quality Institute initiated Care Transitions, an 18-month project to reduce 30-day all-cause readmissions in the Harlingen referral region of the Lower Rio Grande Valley in South Texas. The goal of the project was to engage inpatient hospitals and their “downstream” or discharge providers (e.g., home health agencies, long-term care facilities,

A P S D

Using Rapid Cycle to Implement Preprinted Orders

Use of orders V.4 by all physicians and nurses

Will preprinted orders be useful for acute myocardial infarction patients?

Lea rnin

g Cycle 5: Implement V.4; conduct peer review of documentation and use

Cycle 4: One-week trial of V.3 on the unit

Cycle 3: Two physicians do trial of V.2 for two days

Cycle 2: Dr. A uses V.1 on one patient

Cycle 1: Gather sample orders; have Dr. A provide feedback

EXHIBIT 4.3 Example of

Rapid-Cycle Testing

Note: V.1, V.2, V.3, and V.4 refer to the consecutive versions of the preprinted order sets being tested. Each time the orders are modified during a test, a new version of the orders is created.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

101C h a p t e r 4 : Q u a l i t y I m p r o v e m e n t 101

inpatient rehabilitation facilities) in identifying gaps in care coordination and implementing evidence-based interventions to reduce unnecessary hospital readmissions. As part of the Centers for Medicare & Medicaid Services’ Quality Improvement Organization Program’s 9th Scope of Work, TMF pro- posed that home health agencies, hospices, skilled nursing facilities (SNFs), inpatient rehabilitation facilities (IRFs), and hospitals working in collabora- tion with each other and with physicians could achieve the goals of the Care Transitions project through

• improved communication during the transition of patients from one setting to another,

• use of community and provider-specific data reports to increase accountability and feedback on progress toward goals, and

• implementation of provider-specific evidence-based interventions focused on improving the quality of care during transitions.

During the recruitment phase of the project, TMF engaged 5 inpa- tient hospitals, 28 home health agencies, 11 SNFs, and 2 IRFs. Initial plan- ning at the participating hospitals involved conducting a process-of-care investigation to determine the root causes of their readmission rates. The investigation included the following activities:

• Conducting staff interviews and interdisciplinary meetings to discuss the current discharge process in comparison to Project RED (Re-Engineered Discharge) and to identify barriers and areas for improvement

• Analyzing project data provided by TMF (calendar year 2007 Medicare claims), which included the facility’s 30-day readmission rate and discharge disposition (i.e., home, SNF, IRF, and long-term acute care hospital) in relation to the 30-day readmission rate

• Evaluating current Hospital Consumer Assessment of Healthcare Providers and Systems scores related to the hospital discharge process

The hospitals identified the following root causes (TMF 2010):

• A weak or fragmented discharge plan • Miscommunication or failure to communicate key information at the

time of transition • Discharged patients’ unpreparedness for discharge or self-management • Inadequate medical follow-up with discharged patients after discharge • Inadequate or poor communication with patients and/or caregivers

when relating information about medicines, tests, and red flags

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k102

Following the process-of-care and root cause investigations, the par- ticipating providers reviewed multiple hospital-based interventions designed to reduce unnecessary readmissions, such as (TMF 2010)

• Project RED, • Project BOOST (Better Outcomes for Older adults through Safe

Transitions), • Care Transitions program’s Care Transitions Intervention, and • IHI’s guide to creating an ideal transition home.

Following review of the interventions, all hospitals participating in the Texas Care Transitions project chose to implement components of Project RED. Developed from a study conducted by Boston Medical Center, Project RED includes 11 components targeting patient education, discharge plan- ning, and postdischarge reinforcement:

1. Educate the patient about his or her diagnosis throughout the hospital stay.

2. Make appointments for clinical follow-up visits and testing prior to hospital discharge.

3. Discuss any tests or studies with the patient that have been completed in the hospital, and identify who will be responsible for following up on the results.

4. Organize postdischarge services. 5. Confirm the patient’s medication plan. 6. Reconcile the discharge plan with national guidelines and critical

pathways. 7. Review with the patient the steps he or she should follow if a problem

arises after discharge. 8. Expedite dissemination of the discharge summary to the patient’s

physician and other clinicians involved in the patient’s follow-up care after discharge.

9. Give the patient a written discharge plan at the time of discharge. 10. Implement “teach back” of the patient’s discharge plan by asking the

patient to explain the details of the plan in his or her own words. 11. Follow up on the discharge plan with the patient via telephone two to

three days after discharge.

Throughout the Care Transitions project, TMF provided the follow- ing support to participating providers:

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

103C h a p t e r 4 : Q u a l i t y I m p r o v e m e n t 103

• On-site technical support for team leaders, facility leaders, and Care Transitions committees

• Regional meetings in which community providers could work together across the care continuum to develop region- or community-specific solutions

• Reports identifying the percentage of patients readmitted within 30 days who received a visit from a physician between hospital discharge and readmission

• Quarterly data reports and run charts (based on Medicare claims data) displaying readmission rate performance

• Medical staff education and provider education sessions (e.g., medication reconciliation and health literacy)

• Data collection tools for monitoring the effectiveness of the implemented project components

• A patient discharge survey tool for monitoring the effectiveness of the implemented project components and ensuring that discharge plans met hospital core measurement requirements and national guidelines for patients with acute myocardial infarction, congestive heart failure, or pneumonia

Project results from one of the participating hospitals (see Exhibits 4.4 and 4.5) suggest that the implementation of a community-based project in which providers across the patient care continuum work together can reduce unnecessary hospital readmissions. Support from leadership, accountability for implementation of evidence-based interventions, and concurrent moni- toring are critical to sustaining process redesign efforts. Collaboration among providers across the community on behalf of the patient fosters an awareness of other individual and organizational efforts and successes in overcoming

21.9% 23.1% 22.3% 22.2%

23.7% 23.0% 21.5%

22.6% 22.3%

19.5%

14.0% 16.0% 18.0% 20.0% 22.0% 24.0% 26.0%

CY 2007 Baseline Q2 2008 Q3 2008 Q4 2008 Q1 2009 Q2 2009 Q3 2009 Q4 2009 Q1 2010

VBMC-B Harlingen HRR Target (Q1 2010)

EXHIBIT 4.4 Percentage of 30-Day Readmissions at One Participating Hospital (semiannual rate ending in the listed quarter)

Source: TMF Health Quality Institute. Used with permission.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k104

mutual impediments to improvement. Collective problem solving can expe- dite the application of evidence-based care practices and the use of process redesign methods.

Conclusion

An organization’s success depends on the foundation on which it was built and the strength of the systems, processes, tools, and methods it uses to sustain benchmark levels of performance and to identify and improve per- formance when expectations are not being met. Although quality improve- ment theory and methodology have been available since the early 1900s, their widespread acceptance and application by the healthcare industry have not occurred as rapidly and effectively as in other industries (e.g., manufac- turing). The release of two Institute of Medicine publications (Crossing the Quality Chasm [IOM 2001] and To Err Is Human [Kohn, Corrigan, and Donaldson 2000]) describing significant concerns about the US healthcare system incited a movement toward improvement that greatly increased healthcare institutions’ focus on better care and patient safety (Berwick and Leape 2005). However, because of a combination of technical complexity, system fragmentation, a tradition of autonomy, and hierarchical authority structures, overcoming the “daunting barrier to creating the habits and beliefs of common purpose, teamwork and individual accountability” neces- sary for spread and sustainability will require a continual focus and commit- ment (Berwick and Leape 2005). Sustainable improvement is further defined through will, ideas, and execution. “You have to have the will to improve,

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

HHA Home IRF LTAC SNF Total

Hospital Q1 2008

Hospital Q1 2010

HHRR Q1 2010

EXHIBIT 4.5 Percentage of

Discharges with a 30-Day

Readmission to One

Participating Hospital

Source: TMF Health Quality Institute. Used with permission.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

105C h a p t e r 4 : Q u a l i t y I m p r o v e m e n t 105

you have to have ideas about alternatives to the status quo, and then you have to make it real—execution” (Nolan 2007). The principles described in this chapter have demonstrated success in many healthcare organizations. As healthcare technology advances and access to care improves, healthcare must continue to build on these principles as it strives to reach and maintain benchmark levels of performance. Successful coordination of care across the healthcare continuum will provide the right care for every patient at the right time, every time.

study Questions

1. How would you select and implement one or more of the approaches described in this chapter in your own institution?

2. What are some of the challenges to spreading change? Identify two key questions/issues that need to be considered when applying change concepts in an organization or system.

3. How would a healthcare organization choose elements to measure and measurement tools when seeking to improve the quality of care?

4. How would you encourage your organization to work with other healthcare organizations across the healthcare continuum? Name two factors that are key to ensuring collaboration/coordination among healthcare providers.

5. What are some of the key elements common to the different tools discussed in this chapter?

6. What is the difference between a quality improvement system and a quality improvement tool? Provide examples of each.

references

American Society for Quality (ASQ). 2014. “A. V. Feigenbaum: Laying the Founda- tions of Modern Quality Control.” Accessed January 30. http://asq.org/ about-asq/who-we-are/bio_feigen.html.

American Society for Quality (ASQ) Quality Management Division. 1999. The Certi- fied Quality Manager Handbook. Milwaukee, WI: ASQ Quality Press.

Balanced Scorecard Institute. 2014. “Balanced Scorecard Basics.” Accessed January 30. http://balancedscorecard.org/Resources/AbouttheBalancedScorecard/ tabid/55/Default.aspx.

Baldrige Performance Excellence Program. 2013. 2013–2014 Health Care Criteria for Performance Excellence. Gaithersburg, MD: US Department of Com- merce, National Institute of Standards and Technology.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

http://asq.org/

http://balancedscorecard.org/Resources/AbouttheBalancedScorecard/

T h e H e a l t h c a r e Q u a l i t y B o o k106

Berwick, D. A., and L. L. Leape. 2005. “Five Years After To Err Is Human: What Have We Learned?” Journal of the American Medical Association 293 (19): 2384–90.

Brown, J. A. 2003. The Healthcare Quality Handbook: A Professional Resource and Study Guide. Pasadena, CA: JB Enterprises.

Crosby, P. B. 1996. Quality Is Still Free: Making Quality Certain in Uncertain Times. New York: McGraw-Hill.

Cutler, A. N. 2001. “Biography of Walter A. Shewhart.” www.sigma-engineering. co.uk/ light/shewhartbiog.htm.

Deming, W. E. 2000a. The New Economics for Industry, Government, Education, second edition. Cambridge, MA: MIT Press.

———. 2000b. Out of the Crisis. Cambridge, MA: MIT Press. Feigenbaum, A. V. 1951. Total Quality Control. New York: McGraw-Hill. Heim, K. 1999. “Creating Continuous Improvement Synergy with Lean and TOC.”

Paper presented at the American Society for Quality Annual Quality Con- gress, Anaheim, California, May.

Hertz, H. S. (ed.). 2010. Education Criteria for Performance Excellence (2009– 2010): Baldrige National Quality Program. Darby, PA: DIANE Publishing.

Institute of Medicine (IOM). 2001. Crossing the Quality Chasm: A New Health Sys- tem for the 21st Century. Washington, DC: National Academies Press.

Juran, J. M. 1989. Juran on Leadership for Quality. New York: Free Press. Juran, J. M., and F. M. Gryna (eds.). 1951. Juran’s Quality Control Handbook. New

York: McGraw-Hill. Kilian, C. 1988. The World of W. Edwards Deming. Knoxville, TN: SPC Press. Kohn, L.T., J.M. Corrigan, and M.S. Donaldson (eds.). 2000. To Err Is Human:

Building a Safer Health System. Washington, DC: National Academies Press. Langley, G., K. Nolan, T. Nolan, C. Norman, and L. Provost. 1996. The Improve-

ment Guide: A Practical Approach to Enhancing Organizational Performance. San Francisco: Jossey-Bass.

Massoud, M. R., G. A. Nielson, K. Nolan, T. Nolan, M. W. Schall, and C. Sevin. 2006. “A Framework for Spread: From Local Improvements to System-Wide Change.” IHI Innovation Series white paper. Cambridge, MA: Institute for Healthcare Improvement.

Neave, H. R. 1990. The Deming Dimension. Knoxville, TN: SPC Press. Nolan, T. W. 2007. “Execution of Strategic Improvement Initiatives to Produce

System-Level Results.” IHI Innovation Series white paper. Cambridge, MA: Institute for Healthcare Improvement.

Pande, P. S., R. P. Neuman, and R. R. Cavanagh. 2000. The Six Sigma Way: How GE, Motorola, and Other Top Companies Are Honing Their Performance. New York: McGraw-Hill.

QualityGurus.com. 2014. “Armand V. Feigenbaum.” Accessed January 30. www. qualitygurus.com/gurus/list-of-gurus/armand-v-feigenbaum.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

http://www.sigma-engineering

http://www.qualitygurus.com/gurus/list-of-gurus/armand-v-feigenbaum

107C h a p t e r 4 : Q u a l i t y I m p r o v e m e n t 107

Tague, N. R. 2004. The Quality Toolbox, second edition. Milwaukee, WI: ASQ Qual- ity Press.

Thorndike, E. L., and R. S. Woodworth. 1901. “The Influence of Improvement in One Mental Function upon the Efficiency of Other Functions.” Psychological Review 8: 247–61.

TMF Health Quality Institute (TMF). 2010. Re-Engineering Discharges in a Com- munity-wide Project Reduces 30-Day Hospital Readmission Rate SQUIRE. Austin, TX: TMF Health Quality Institute.

Womack, J. P., and D. T. Jones. 2003. Lean Thinking: Banish Waste and Create Wealth in Your Corporation. New York: Free Press.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

Co py ri gh t © 2 01 4. H ea

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

PART

II HeaLtHCare QuaLity at tHe organization anD

MiCrosysteM LeVeLs

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

CHAPTER

111

5Data CoLLeCtion John J. Byrnes

e verywhere you turn, everyone wants data. What do they really mean? Where do you get data? Is chart review the gold standard, the best source? Are administrative databases reliable; can they be the gold stan-

dard? What about health plan claims databases—are they accurate? What is the best source for inpatient data that reflects the quality of patient care from both a process and an outcome perspective? When working in the outpatient environment, where and how would you obtain data that reflect the level of quality delivered in physician office practices? These questions challenge many healthcare leaders as they struggle to develop quality improvement and measurement programs. This chapter clarifies these issues and common industry myths and provides a practical framework for obtaining valid, accu- rate, and useful data for quality improvement work.

Categories of Data: Case example

Quality measurements can be grouped into four categories or domains: (1) clinical quality (including both process and outcome measures); (2) financial performance; (3) patient, physician, and staff satisfaction; and (4) functional status. To report on each of these categories, one may need to collect data from several separate sources. The challenge is to collect as many data ele- ments from as few data sources as possible with the objectives of consistency and continuity in mind. For most large and mature quality improvement projects, teams will want to report their organization’s performance in all four domains.

Spectrum Health’s clinical reporting (CR) system illustrates this point. The CR system contains more than 50 disease-specific dashboards that report performance at the system, hospital, and physician levels (see Exhibit 5.1). In Exhibit 5.2, a dashboard for total hip replacement provides examples of clini- cal quality and financial performance measures. To produce the CR system, Spectrum Health used a variety of data sources, including extracts from its finance and electronic health record (EHR) systems. The decision support department processed the data, applying a series of rigorous data cleanup

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k112

algorithms, adjusting for severity, and adding industry benchmarks. The resulting report contains measures of clinical processes (antibiotic utiliza- tion, deep vein thrombosis [DVT] prophylaxis, beta-blocker administration, autologous blood collection, and blood product administration), financial performance (lengths of stay, total patient charges, pharmacy charges, lab charges, X-ray charges, and intravenous therapy charges), and clinical out- comes (DVT, acute myocardial infarction [AMI], and readmission within 31 days). From more than 200 indicators available in the database, the total joint quality improvement team selected these measures as the most important for assessing the quality and cost of care delivered. The measures also include some Joint Commission core measures.1

To obtain patient satisfaction information, the team uses industry- standard patient satisfaction surveys. The outbound call center administers these surveys by telephone within one week of a patient’s discharge. The results can be reported by nursing unit or physician, are updated monthly, and can be charted over the past eight quarters.

1. Chest pain 2. Heart attack 3. PCI 4. Heart failure 5. Pneumonia 6. Normal delivery 7. C-section 8. Bypass surgery 9. Valve surgery

10. Stroke—ischemic 11. Total hip replacement 12. Total knee replacement 13. Hip fracture 14. Abd. hysterectomy—non-CA 15. Abd. hysterectomy—CA 16. Lap hysterectomy 17. Cholecystectomy—lap 18. Cholecystectomy—open 19. Lumbar fusion 20. Lumbar laminectomy 21. Bariatric surgery 22. Colon resection 23. Diabetes and glycemic control 24. DVT 25. COPD 26. Upper GI bleed 27. SCIP 28. Peripheral vascular procedures

29. Pediatric asthma 30. Very low birth weight neonates 31. Pediatric appendectomy 32. RSV/bronchiolitis 33. Pediatric chemotherapy 34. Pediatric VP shunts 35. Pediatric hospitalist conditions

a. Bronchitis and asthma b. Esophagitis and

gastroenteritis c. Kidney and UTI d. Nutritional and miscellaneous

metabolic disorders e. Otitis media and URI f. Pediatric pneumonia

g. Seizure and headache h. Fever of unknown origin

36. NICU, PICU, and adult ICU (medi- cal, surgical, and burn)

37. AHRQ patient safety indicators 38. Pain management 39. Sickle cell 40. Sepsis 41. 100,000 Lives Campaign 42. 5 Million Lives Campaign 43. National Patient Safety Goals 44. Rapid response team

EXHIBIT 5.1 Spectrum

Health’s Clinical Reporting System— Available

Disease and Project Reports

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

C h a p t e r 5 : D a t a C o l l e c t i o n 113

To complete the measurement set, the team includes the results of patients’ functional status (following their treatments). This information can be obtained from patients’ EHRs (if it has been included in them) or by using survey tools during follow-up visits. Many hospital procedures are performed to improve patients’ functional status. A patient who undergoes a total knee replacement, for example, should experience less knee pain when he or she walks, have a good range of joint motion, and be able to perform the activities of daily living that most of us take for granted. For this report, the team examines patients’ functional status before and after hospitalization to demonstrate that their treatments were effective.

In summary, when designing data collection efforts, quality improve- ment teams need to maintain a balanced perspective of the process of care by collecting data in all four categories: clinical quality, financial performance,

Administrative Data Process Coumadin Blood No. of 1st gen. Low mol. or LMW Beta Autologous prod. DVT Hip Name patients Ceph Vancomycin Coumadin Heparin wt. heparin heparin blocker blood coll. given prophylaxis* revision

BL 617 95.5% 9.9% 14.6% 23.0% 91.2% 96.6% 39.9% 1.8% 33.2% 99.7% 20.4% BW 136 90.4% 11.8% 5.9% 5.1% 100.0% 100.0% 41.9% 4.4% 30.9% 100.0% 13.2% SH-GR 753 94.6% 10.2% 13.0% 19.8% 92.8% 97.2% 40.2% 2.3% 32.8% 99.7% 19.1%

Administrative Data Outcome Education Any Education No. of readmit AMI participation Name patients DVT AccPuncLac 30 days 2nd DX Los rate*

BL 617 0.6% 0.0% 4.2% 0.0% 3.67 59.3% BW 136 0.0% 0.0% 4.4% 0.7% 3.78

** The education rate reflects all total joint replacement patients who had their SH-GR 753 0.5% 0.0% 4.2% 0.1% 3.69 surgery within the time period stated on this dashboard.

JCAHO SCIP JCAHO Surgical Care Improvement Project No. of Preop dose Antibiotic Selection Postop duration Name patients (SCIP-INF-1)* (SCIP-INF-2) (SCIP-INF-3)*

SH-GR Varies 96.0% n = 75 100.0% n = 76 97.2% n = 72

Administrative Data Direct Costs No. of ICU Laboratory OR Pharmacy Radiology R&B Supplies Therapy Other Total Name patients cost cost cost cost cost cost cost cost cost cost

BL 617 $71 $180 $2,219 $384 $79 $1,460 $1,944 $394 $217 $6,948 BW 136 $101 $127 $1,140 $405 $101 $1,801 $5,062 $389 $285 $9,410 SH-GR 753 $76 $170 $2,024 $388 $83 $1,521 $2,507 $393 $230 $7,393

Administrative Data Fully Allocated Costs No. of ICU Laboratory OR Pharmacy Radiology R&B Supplies Therapy Other Total Name patients cost cost cost cost cost cost cost cost cost cost

BL 617 $117 $251 $3,711 $492 $162 $3,020 $2,078 $559 $326 $10,715 BW 136 $189 $176 $2,279 $515 $171 $3,215 $5,263 $578 $416 $12,802 SH-GR 753 $130 $237 $3,452 $496 $163 $3,055 $2,653 $562 $342 $11,092

Administrative Data Potential Direct Cost Savings No. of Total cost Name patients DVT AccPuncLac AMI 2nd DX (Patients above average)

BL Varies $51,618 n = 4 $0 n = 0 $0 n = 0 $679,916 n = 189 BW Varies $0 n = 0 $0 n = 0 $9,653 n = 1 $165,825 n = 61 SH-GR Varies $49,770 n = 4 $0 n = 0 $11,614 n = 1 $920,655 n = 270

* Denotes indicators selected for “The Joint Commission”

Prepared June 10, 2007 by the Spectrum Health Quality Department.

Spectrum Health Clinical Outcomes Report (COR)–Hip Replacement March 1, 2006 to February 28, 2007

EXHIBIT 5.2 Clinical Dash- board—Hip Replacement

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k114

patient satisfaction, and functional status. Teams that fail to maintain this bal- ance may overlook critical information. For instance, a health system in the Southwest initially reported that it had completed a series of successful quality improvement projects—clinical care had improved, patient satisfaction was at an all-time high, and patient outcomes were at national benchmark levels. However, subsequent review of the projects identified that some of the inter- ventions had negatively affected the system’s financial outcomes. Revenue had significantly decreased as a result of several interventions, and other interven- tions had increased the cost of care. If financial measures had been included in the reporting process, the negative financial effect could have been minimized and the same outstanding quality improvements would have resulted. In the end, the projects were considered only marginally successful because they lacked a balanced approach to process improvement and measurement.

Considerations in Data Collection

Time and Cost Involved in Data Collection All data collection efforts take time and money. The key is to balance the cost of data collection and the value of the data to your improvement efforts. In other words, are the cost and time spent collecting data worth the effort? Will the data have the power to drive change and improvement? Although this cost–benefit analysis may not be as tangible as it is in the world of busi- ness and finance, the value equation must be considered. Generally, medi- cal record review and prospective data collection are considered the most time-intensive and expensive ways to collect information. Many reserve these methods for highly specialized improvement projects or use them to answer questions that have surfaced following review of administrative data sets. Use of administrative data2 is often considered cost-effective, especially because the credibility of administrative databases has improved and continues to improve through the efforts of coding and billing regulations, initiatives,3 and rule-based software development. Additionally, third-party vendors can provide data cleanup and severity adjustment. Successful data collection strategies often combine both code- and chart-based sources into a data col- lection plan that capitalizes on the strengths and cost-effectiveness of each.

The following situation illustrates how the cost-effectiveness of an administrative system can be combined with the detailed information available in a medical record review. A data analyst using a clinical decision support sys- tem (administrative database) discovered a higher-than-expected incidence of renal failure (a serious complication) following coronary artery bypass surgery. The rate was well above 10 percent for the most recent 12 months (more than 800 patients were included in the data set) and had slowly increased over the past six quarters. However, the clinical decision support system did not

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

C h a p t e r 5 : D a t a C o l l e c t i o n 115

contain enough detail to explain why such a large number of patients were experiencing this complication—whether this complication resulted from the coronary artery bypass graft procedure or was a chronic condition present on admission. To find the answer, the data analyst used chart review to (1) verify that the rate of renal failure as reported in the administrative data sys- tem was correct, (2) isolate cases of postoperative incidence, (3) identify the root cause(s) of the renal failure, and (4) answer additional questions about the patient population that were of interest to the physicians involved in the patients’ care. In this example, the analyst used the administrative system to identify unwanted complications in a large patient population (a screening or surveillance function) and reserved chart review for a much smaller focused study (80 charts) to validate the incidence and determine why the patients were experiencing the complication. This excellent example shows effective use of two common data sources and demonstrates how the analyst is able to capitalize on the strengths of both while using each most efficiently.

Collecting the Critical Few Rather than Collecting for a Rainy Day Many quality improvement efforts collect every possible data element in case it might be needed. Ironically, justification for this approach is often based on saving time—the chart has already been pulled, so one might as well be thorough. This syndrome of stockpiling “just in case” versus fulfilling requirements “just in time” has been studied in supply chain management and proven to be ineffective and inefficient. It also creates quality issues (Denison 2002). This approach provides little value to the data collection effort and is one of the biggest mistakes quality improvement teams make. Rather than provide a rich source of information, this approach unnecessarily drives up the cost of data collection, slows the data collection process, creates data management issues, and overwhelms the quality improvement team with too much information.

For all quality improvement projects, it is critical to collect only the data required to identify and correct quality issues. As a rule in ongoing data collection efforts, quality improvement teams should be able to link every data element collected to a report, thereby ensuring that teams do not collect data that will not be used (James 2003). In the reporting project discussed earlier, the hospital team was limited to selecting no more than 15 measures for each clinical condition. It also selected indicators that (1) have been shown by evidence-based literature to have the greatest effect on patient outcomes (e.g., in congestive heart failure, the use of angiotensin converting enzyme [ACE] inhibitors and beta blockers and evaluation of left ventricular ejection fraction); (2) reflect areas in which significant improvements are needed; (3) will be reported in the public domain (Joint Commission core measures); and (4) together provide a balanced view of the clinical process of care, financial performance, and patient outcomes.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k116

Inpatient Versus Outpatient Data The distinction between inpatient and outpatient data is an important con- sideration in planning the data collection process because the data sources and approaches to data collection can be different.

The case of a team working on a diabetes disease management project illustrates this point. First, disease management projects tend to focus on the entire continuum of care, so the team needs data from both inpatient and outpatient settings. Second, the team needs to identify whether patients receive the majority of care in one setting or the other and decide whether data collection priorities should be established with this setting in mind. For diabetes, the outpatient setting has the most influence on patient outcomes, so collection of outpatient data is a priority. Third, the team must select the measures that reflect the aspects of care that have the most influence on patient outcomes. Remembering to collect the critical few (as discussed in the previous section), the team would consult the American Diabetes Association (ADA) guidelines for expert direction. Fourth, the team must recognize that the sources of outpatient data are much different from the sources of inpa- tient data, and outpatient data tend to be more fragmented and harder to obtain. However, with the advent of outpatient EHRs and patient registries, the ease of collecting outpatient data is improving.

To identify outpatient data sources, the team should consider the fol- lowing questions:

• Are the physicians in organized medical groups that have outpatient EHRs? Can their financial or billing systems identify all patients with diabetes in their practices? If not, can the health plans in the area supply the data by practice site or individual physician?

• Some of the most important diabetes measures are based on laboratory testing. Do the physicians have their own labs? If so, do they archive the lab data for a 12- to 24-month snapshot? If they do not do their own lab testing, do they use a common reference lab that would be able to supply the data?

Once the team answers these questions, it will be ready to proceed with data collection in the outpatient setting.

sources of Data

As just discussed, the sources of data for quality improvement projects are extensive. Some sources are simple to access, while accessing others is com- plex; some data sources are inexpensive to use, while others are expensive. In

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

C h a p t e r 5 : D a t a C o l l e c t i o n 117

the average hospital or health system, data sources include medical records, prospective data collection, surveys of various types, telephone interviews, focus groups, administrative databases, health plan claims databases, cost accounting systems, patient registries, stand-alone clinical databases, EHRs, and lab and pharmacy databases.

The following objectives are essential to a successful quality improve- ment project and data collection initiative:

• Identify the purpose of the data measurement activity (i.e., for monitoring at regular intervals, investigation over a limited period, or a onetime study).

• Identify data sources that are most appropriate for the activity. • Identify the most important measures to collect (the critical few). • Design a common-sense strategy that will ensure collection of

complete, accurate, and timely information.

By following these steps, project teams will gather actionable data and the information required to drive quality improvements.

Medical Record Review (Retrospective) Retrospective data collection involves identification and selection of a patient’s medical record or group of records after the patient has been discharged from the hospital or clinic. Records generally cannot be reviewed until all medical and financial coding is complete because codes are used as a starting point for identifying the study cohort.

For several reasons, many quality improvement projects depend on medical record review for data collection. First, many proponents of medical record review believe it to be the most accurate method of data collection. They believe that because administrative databases have been designed for financial and administrative purposes rather than for quality improvement, the databases contain inadequate detail, many errors, and “dirty data”—that is, data that make no sense or appear to have come from other sources.

Second, some improvement projects rely on medical record review because many of the data elements are not available from administrative data- bases. For example, most administrative databases do not contain measures that require a time stamp, such as administration of antibiotics within one hour before surgical incision.

Third, several national quality improvement database projects— including the Healthcare Effectiveness Data and Information Set (HEDIS), Joint Commission core measures, Leapfrog Hospital Survey,4 and National Quality Forum’s (NQF) National Voluntary Consensus Standards for Hos- pital Care—depend on retrospective medical record review for collecting a

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

T h e H e a l t h c a r e Q u a l i t y B o o k118

significant portion of the data elements required to be reported. The records not only contain measures requiring a time stamp but, for some measures, also require the data collector to include or exclude patients on the basis of criteria that administrative databases do not capture consistently. The per- centage of patients with congestive heart failure who are receiving an ACE inhibitor is an example of this type of measure. The use of ACE inhibitors in this population is indicated for all patients with an ejection fraction of less than 40 percent. The ejection fraction is not part of the typical administrative database. Sometimes this information is contained in a generally inaccessible, stand-alone database in the cardiology department, or it may be contained only in a transcribed report in the patient’s medical record. Hence, accurate reporting of this measure, one of the most critical interventions that a patient with congestive heart failure will receive, depends completely on retrospec- tive chart review. A consensus document presented to NQF5 suggested that clinical importance should rate foremost among criteria for effectiveness and that measures that score poorly on feasibility6 because of the burden of medi- cal record review should not be excluded solely on that basis if their clini- cal importance is high (NQF Consumer, Purchaser, and Research Council Members 2002).

Fourth, focused medical record review is the primary tool for answer- ing the “why” of given situations (e.g., why patients were experiencing a particular complication, why a certain intervention negatively affected patient outcomes). Medical record review continues to be a key component of many data collection projects, but it needs to be used judiciously because of the time and cost involved.

The approach to medical record review involves a series of well- conceived steps, beginning with the development of a data collection tool and ending with the compilation of collected data elements into a registry or electronic database for review and analysis.

Prospective Data Collection, Data Collection Forms, and Scanners Prospective data collection also relies on medical record review, but it is com- pleted during a patient’s hospitalization or visit rather than retrospectively. Nursing staff, dedicated research assistants, or full-time data analysts com- monly collect the data. The downside to asking nursing staff to collect data is the effort involved; it is a time-consuming task that can distract nurses from their direct patient care responsibilities. A better approach would be to hire research assistants or full-time data analysts who can collect the data and be responsible for data entry and analysis. Because this job is their sole respon- sibility, the accuracy of data collection is greater. If they also are responsible for presenting their analyses to various quality committees, they are likely to review the data more rigorously.

Copying and distribution of this PDF is prohibited without written permission. For permission, please contact Copyright Clearance Center at www.copyright.com

lt h Ad mi ni st ra ti on P re ss . Al l ri gh ts r es er ve d. M ay n ot b e re pr od uc ed i n an y fo rm w it ho ut p er mi ss io n fr om t he p ub li sh er , ex ce pt f ai r us es p er mi tt ed

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .