Using IT to Improve Quality at NewYork-Presybterian Hospital: A Requirements-Driven Strategic Planning Process

Abstract

At NewYork-Presbyterian Hospital, we are committed to the delivery of high quality care. We have implemented a strategic planning process to determine the information technology initiatives that will best help us improve quality. The process began with the creation of a Clinical Quality and IT Committee. The Committee identified 2 high priority goals that would enable demonstrably high quality care: 1) excellence at data warehousing, and 2) optimal use of automated clinical documentation to capture encounter-related quality and safety data. For each high priority goal, a working group was created to develop specific recommendations. The Data Warehousing subgroup has recommended the implementation of an architecture management process and an improved ability for users to get access to aggregate data. The Structured Documentation subgroup is establishing recommendations for a documentation template creation process. The strategic planning process at times is slow, but assures that the organization is focusing on the information technology activities most likely to lead to improved quality.

INTRODUCTION

Health care is in the midst of an upheaval.^{1, 2} The increase in health care costs continues to outpace inflation at the same time that abundant evidence makes clear that the safety and quality of healthcare are lacking.^{3, 4, 5} New and innovative approaches to health care quality improvement and cost reduction are continually being sought.^{6, 7, 8, 9} In short, there is a pervasive concern that the current health care system is unsustainable.

Whatever will be an eventual sustainable model of health care, there is virtually universal agreement that the system will include a level of health information technology (IT) far beyond what we have today. One vision of an eventual health care infrastructure includes “delivering information to individuals – consumers, patients, and professionals – when and where they need it”, so the information can be used in decision making.¹⁰ Clearly, as a society we have a long way to go to achieve this goal.

Health care organizations that are on the forefront of the health care delivery system have a key role to play in implementing society’s vision of the way IT will be brought to bear on the problems of the health care system.⁷ Such organizations are responsible for choosing the technologies that are best suited to addressing the quality problems at hand and for implementing them in a way that allows potential benefits to be realized. In addition to selecting technologies that solve individual problems, organizations must assure that the overall information technology architecture is coherent and well supported and supports the business goals of the institution.¹¹

Health care organizations must overcome significant challenges as they set about to achieve these goals. One of the most formidable challenges for a health care organization is the creation of a strategy that will allow it to achieve its quality goals. Glaser¹², ^p2–3 has pointed out that the strategy of an organization includes formulation (i.e., making decisions about the mission and goals of the organization and the activities and initiatives that it will undertake to achieve those goals) and implementation (decisions about the structures, staff, and processes that will be put into place to achieve the goals). Glaser also notes ¹², ^p8 that an IS department’s strategy should be linked to the larger organization’s overall strategy. IT strategies often are derived from organizational strategies. This can be challenging when the larger organization’s strategy is not explicit or is in a state of flux. At times, an IT strategy may create new strategies for the overall organization (e.g., when a retail organization extends its reach via the Web).

Besides strategy, another important concept in IT is that of requirements definition. ¹³, ^p29Requirements articulate to the technical staff what the information technology is supposed to achieve. The importance of requirements is highlighted by the observation that the cost of implementing a change to software at the requirements stage is 50–200 times less than implementing the same change in the software construction or maintenance stage.¹³, ^p25 Explicit requirements are especially important in the domain of health IT, where the complexities of clinical medicine must be managed in the software as flawlessly as possible.¹⁴

At NewYork-Presbyterian Hospital, we have undertaken a planning process to help us decide how best to use information technology to advance our quality goals. Our approach has taken into account the importance and complexity of strategic planning¹² as well as a desire to create explicit requirements for the Information Systems department. We began the process about 18 months ago. This paper reports on our progress to date.

SETTING AND CONTEXT

NewYork-Presbyterian Hospital (NYPH) is the largest hospital in New York City. It is a multi-campus hospital, with about 2400 beds and $2.2 billion dollars in annual revenue. NYPH was created in 1999 from the merger of Columbia-Presbyterian Medical Center and The New York Hospital. NYPH is affiliated with the medical schools of Columbia University and Cornell University. Faculty physicians are affiliated with one of the two medical schools. NYPH’s 5 strategic goals are: 1) Staff development, 2) Providing the highest quality care, 3) Identifying innovative approaches to the complexities of health care, 4) Growing the volume of business that we deliver, and 5) Using information technology to its fullest capacity.

In August 2004, the NYPH Chief Medical Officer (CMO) commissioned an initiative to determine how we should be using information technology to improve the quality of care we deliver. In chartering the effort, the CMO noted that we need to be confident that we are delivering the highest quality care to our patients. Such forces as consumer driven health care, pay for performance, and continuing challenges with risk management reinforce the need to deliver demonstrably excellent care. He noted that for many measures, data collection and analysis is manual, untimely, and expensive. He noted that, ideally, data from across the organization for a wide variety of process and outcome measures would be available in a timelier manner and there would be coordination of inpatient and outpatient data. He noted that although we were using information technology in several ways to improve quality – for example, we were implementing computerized physician order entry (CPOE) and ambulatory electronic health record (EHR) systems and we had a number of data warehouses we used for generating reports for regulatory purposes, etc. -- we did not have an explicit strategic approach to the use of information systems to improve quality. The charge from the CMO was to develop a strategic plan for the use of IT to improve quality.

CLINICAL QUALITY AND IT (CQIT) COMMITTEE

To respond to the CMO’s request, NYPH formed a Clinical Quality and IT (CQIT) Committee. The mission of the CQIT Committee was to make recommendations to senior leadership regarding ways that information technology can be used to improve quality. The CQIT Committee had 30 members including physicians from both medical schools and representatives from such hospital departments as the Laboratory, Pharmacy, Nursing, Information Systems, and Quality. The Committee also included senior executives, such as the Vice-President for Medical Affairs and the Senior Vice President for Strategic Planning. Scheduling the committee was a formidable task, however broad multi-disciplinary representation was felt to be essential for the initiative to be successful. The first meeting in December 2004 affirmed that there was broad support for the initiative within the Hospital.

In March 2005, the CQIT Committee reviewed the Hospital’s quality vision. (A common understanding of the organization’s quality vision was deemed to be essential if the Committee was going to make meaningful recommendations for how best to use IT to improve quality.) The NYPH quality agenda includes:

1. Addressing regulatory reporting requirements. NYPH reports performance measures to over a dozen organizations including the Center for Medicare and Medicaid Services (CMS), the NY State Department of Health, and the National Surgical Quality Improvement program.

2. Patient safety. NYPH has programs to address the Joint Commission’s National Patient Safety goals¹⁵ as well as internally motivated safety initiatives.¹⁶ NYPH’s overall safety efforts include risk reduction and infection control.

3. Pay for performance. NYPH increasingly must demonstrate compliance with quality measures that are part of pay for performance contracts.⁸ These measures may be ad hoc, or may overlap with measures created by national agencies (e.g., CMS).

4. Performance excellence. NYPH’s internal quality improvement program is based on the Six Sigma methodology.¹⁶ Six Sigma is a quantitatively rigorous, measurement-based, quality improvement strategy that focuses on process improvement and variation reduction. Six Sigma has been used effectively in several industries and increasingly is being used in health care.

After NYPH’s dimensions of quality were outlined, the Hospital’s clinical IT plan was reviewed. For the inpatient environment, one campus has a legacy Eclipsys application with CPOE, nurse documentation (notes and medication administration record) and some physician documentation. The other campus is implementing a new generation Eclipsys product that provides the same functions. The new generation Eclipsys product eventually will be deployed house wide. Hospital-owned and faculty ambulatory practices will have a mix of Eclipsys and other vendor EHR systems.

After the quality and the IT plan were laid out, the members of the Committee were led through a facilitated exercise to identify next steps. Starting with each dimension of quality (i.e., regulatory, safety, pay for performance, etc.), Committee members contributed individual ideas regarding ways that IT could be used to improve quality. Ideas were then grouped into broad categories. Multi-voting was used to prioritize the categories. The highest priority categories were: being able to track specific safety measures, access to data for analysis and reporting and research, ability to integrate data across applications for reporting purposes, and ability to extract data from clinical notes for quality purposes.

The Committee considered what specific initiatives could help address the high priority categories. Two general themes arose. The first theme was the recognition that, to be demonstrably excellent at quality, the organization would need to have easy access to automated data for performance measurement and other kinds of reporting. The group articulated this requirement by saying the organization would need to be excellent at “data warehousing”.

The second theme that arose was a recognition that as the organization moved from paper-based encounter documentation to automated encounter documentation, it had an opportunity to capture clinical observation data that could be used to measure and improve quality and safety. The group acknowledged that such benefits would only be realized if clinical observations were documented reliably, reproducibly, consistently and in a coded manner. The group named this theme “structured documentation”.

Subgroups were chartered for Data Warehousing theme and the Structured Documentation theme. The charge to the subgroups was to come back with specific recommendations by year’s end. The CQIT Committee continued to meet bimonthly as a steering committee

CQIT SUBGROUPS

Data Warehousing subgroup

The Data Warehousing subgroup began its work by reviewing the constituencies at the organization that make use of warehoused data, in other words, the groups that need aggregate data. These constituencies are shown in Table 1.

Table 1.

"Consumers" of warehoused data

1	Regulatory/indicators (repeated reports)
2	Documentation improvement (lists of patients with specific diagnoses, with relevant data)
3	Performance excellence / Six Sigma projects (ad hoc requests for data)
4	Pay for performance activities (reports)
5	Academic researchers (ad hoc queries)
6	Workflow (creation of lists, e.g. for infection control)

After the user needs were well understood, the current warehousing platforms in use at NYPH were studied. These platforms are shown in Table 2.

Table 2.

Warehouses in use at NYPH

1	Clinical Data Warehouse (stores ancillary and encounter data from inpatient environment and ambulatory environment at one campus)
2	Cornell Physicians Organization data warehouse (stores ambulatory encounter and clinical data from the other campus)
3	Transition Systems, Inc. (TSI) warehouse(financial, registration data)
4	Business Intelligence Systems (OLAP environment for selected data types, mostly administrative in nature, e.g., length of stay)
5	Eclipsys warehouse (orders and documentation data)

Information about user needs and the current platforms was obtained via interviews of appropriate representatives.

Once the users’ needs were known, and the current state was known, the Data Warehousing Subgroup did a gap analysis asking: What did the users need that wasn’t being delivered?

The gap analysis yielded two high priority findings. First, there was not a robust process at the organization for users to request data. When someone in the organization has a new data need, they must “hunt around” for the appropriate person to help them. Finding the right group may take some time. Addressing a data need that involves more than one data source (e.g., TSI and Eclipsys) may be especially complex. The group recommended that a more robust way to meet end-user requests for data would be worthwhile and that more analysis be done to identify specific actionable recommendations.

The Data Warehousing Subgroup’s second finding was that addressing needs that spanned data platforms often was complex. For example, there was no easy way to compare medications in the Cornell PO data warehouse (ambulatory setting) with medications in the Eclipsys warehouse (inpatient) because the two medication vocabularies were different. Also, different systems would implement nuances of “standards” in different ways. For example, one laboratory system would extend LOINC in one way and another laboratory system would extend LOINC in a different way. The Data Warehousing subgroup recommended that a more robust process be established to manage the systems architecture, with a focus on vocabulary.

Data Warehousing Subgroup – Current Status

An Architecture Committee was chartered in late 2005 with a mission to assure that the warehouses and departmental systems at NYPH can interoperate in a way that supports the organization’s quality mission. As of this writing, the Architecture Committee is working to identify its specific scope. A Vocabulary Committee has been established to determine what should be the vocabulary architecture that will allow us to realize our quality goals.

Also, a small task force is working to identify how a more robust “data request function” might be organized. A prerequisite to making specific recommendations is obtaining detailed knowledge about the owners, custodians and “power users” of the existing key data systems. The task force plans to have specific recommendations, which would involve resource as well as process specifications, by mid-2006.

Structured Documentation Subgroup

The Structured Documentation Subgroup began its work by identifying the quality-related goals of automated documentation. There were 11 specific goals that fell into 3 categories.

Five success factors were identified that would be necessary for the organization to realize the goals of automated documentation. These are shown in Table 4.

Table 4.

Dependencies for realizing quality with automated documentation

1	Data Warehousing (needed to extract data for quality purposes)
2	Terminology Services (needed to identify the coded concepts in documentation templates)
3	Clinical Leadership and Staffing (needed for documentation template content)
4	Template Engineering Process (needed to assure consistency)
5	Evaluation Tools (measurement, to determine if templates are impacting quality and safety)

The Structured Documentation Subgroup also did an analysis to explain to senior leadership how improved capture of complications and comorbidities could lead to increased hospital revenue as well as increased severity of illness and risk of mortality measures that are used in benchmarking activities, such as the US News and World Reports rankings.

To be able to make robust recommendations regarding the resources the organization should commit to supporting its structured documentation efforts, the subgroup participated actively in 2 pilot automated documentation projects in early 2006 – one in Cardiology and one in Surgery. The Cardiology pilot had the goal of creating a structured automated discharge summary note and the Surgery pilot goal was to create a generalized surgical progress note. Lessons from the pilots were: 1) template creation took longer than expected, 2) a naming convention for documents and a style guide for template creation would be critical, 3) the creation of expandable acronyms (macros) would need to be managed, 4) regulatory compliance and billing considerations would need to be taken into account as part of template creation, and 5) the overall template creation process would need active governance.

Structured Documentation Subgroup – Current Status

The Structured Documentation Subgroup is forging a vision for the management of the template creation process at NYPH. The vision includes a governance model with domain-specific working groups that have expertise in the clinical domain, vocabulary management, billing and compliance, and template design. The working groups would be overseen by a Documentation Steering Committee that would report to the Hospital’s Information Systems Clinical Advisory Group. The resources necessary to support this model (apart from the technical resources) would include a physician lead, someone to manage the activities of the workgroups, and an end-user educational specialist. Vocabulary support also would be needed. Not all these roles need be full time. As of this writing, the Structured Documentation Subcommittee is refining this vision and preparing to submit it for review to senior leadership.

SUMMARY

We are 18 months into this strategic planning process. Progress was slow at first, but has been steady and more substantial over time. The process has required the creation of a large number of committees which at times has seemed bureaucratic, but has fostered broad support for the eventual recommendations. Much energy was dedicated to setting realistic expectations about what could be achieved with information systems and what would be required (both in terms of process and resources) to achieve such goals.

The process has been successful thus far in that we have achieved broad acceptance of what are the themes that the organization needs to focus on (warehousing and automated documentation) if it is to make best use of information technology to improve quality. The focus on quality has been critical to the eventual decisions. For example, some stakeholders were eager to expedite automated documentation by minimizing the governance of the template creation process. It was pointed out that although this approach might in fact increase speed of deployment and end-user acceptance, the organization would not realize any quality benefits.

We realize we are at the beginning of a multi-year process. We are confident that we have created a framework that will allow the organization to make informed decisions going forward about how best to use information systems to improve quality.

Table 3.

Ways automated documentation improves quality

1	Improve clinical care by:
a	increasing compliance with clinical guidelines
b	enabling automated clinical decision support
c	facilitating communication between providers across transitions in care
d	supporting interdisciplinary communication and coordination of care
e	supporting documentation that clearly and accurately reflects care that was delivered
f	supporting intuitive and efficient documentation practices
2	Enhance revenue recovery by:
a	supporting our requirements for indicators
b	supporting “pay-for-performance” contracting
c	supporting documentation improvement efforts
d	supporting automation of billing procedures
3	Support quality and patient safety by:
a	enabling enterprise-wide measurement of quality
b	enabling patient safety measurement activities