The journal Healthcare Quarterly published a special issue in 2006 1Source. I don’t think it’s accessible without a subscription. about the implementation of computerized provider order entry (CPOE) at the University Health Network in Toronto. Below are my notes about the papers in this issue.
Transforming Healthcare Organizations – Brian Golden
- “healthcare organizations are the most complex form of human organization we have ever attempted to manage.” (Drucker, cited in Golden)
- “confluence of [different types of] professionals and other stakeholders (e.g,. patients, government) often with seemingly incompatible interests, perspectives, and time horizons” (Golden)
- there is an “insatiable demand for healthcare – without unlimited financial support”
- medication order entry (MOE)/medication administration record (MAR) is intended to:
- “decrease adverse events related to medication error”
- “provide a rich data set to monitor and improve operational performance”
- challenges related to change management
- disparate stakeholder groups
- multiple organizational missions (e.g., provide care, be fiscally solvent, employ people)
- professional autonomy highly valued by clinicians, who make decisions that “influence a major portion of healthcare expenditures”
- “information needed to managed the change process lacking”
- stages of the change process:
- determine desired end state
- the “vision” or “where we want to be” – needs to be actionable
- at UHN, had “measurable goals [to] allow [UHN] to chart progress and determine success or failure” – “critical to focusing the attention of change leaders and those will be asked (or required) to change”
- assess readiness for change
- “broad situational analysis”, including key stakeholder analysis, what new capabilities will be needed, site’s experience with previous (good or bad) or concurrent (competing) changes
- broaden support & organizational redesign
- communications:
- consistent message about the objectives of the change and how the change will be done
- multiple media for communicating these messages
- messages tailored to audiences
- messages delivered by credible source
- management not talking about an issue until they have all the answer -> rumours and decreased credibility of the change leaders (as people believed they were keeping secrets). Better to be honest about the unknowns and explain your plan to decrease the uncertainty (e.g., say when you expect to know or how you plan to approach things once you have the data you need to make a decision)
- recognize the most “saleable qualities” of the change. i.e., people more likely to accept changes “to they are seen to be:
- “trialed” and revisable
- divisible (phased implementation)
- concrete
- familiar
- congruent
- marginal
- staff need to appreciate the benefits of the change
- organizational environment needs to support the change
- changing one of the points on the star may require changes on another point to realign things
- you can’t change culture and values directly – only indirectly by changing things about the points of the star
- people need not only to be able to implement the change, but also be motivated to implement the change – they have to see the benefit
- benefits can be to patients, the organization, the health system, clinicians, and/or administrations
- communications:
- reinforce and sustain change
- monitoring performance and showcasing successes can help reinforce and sustain the project
- those monitoring/evaluation data can also be used to fine-tune the system
- also need to reward supporters and recognize losses
- determine desired end state
- recipe for a successful change:
- vision + skills + incentive + resources + action plan
The Business Case for Patient Safety – Anderson et al
- Canadian Adverse Events Study (data from 2000 fiscal year from chart audits at hospitals in 5 provinces):
- errors occur in 7.55 of annual hospital admissions
- 36.9% of those errors were “entirely preventable”
- 20.8% of those resulted in the death of a patient
- administration of drugs was “most critical problem contributing to adverse events”
- problem can be from any mix of:
- illegible written orders
- provider writing the wrong prescription
- prescribing the wrong meds
- transcription errors/misinterpreted orders by nurses
- incorrect administration of meds
- incorrect documentation of meds
- CPOE reduces:
- length of stay
- repeat tests
- turnaround times for lab/pharmacy/radiology requests
- costs (Birkmeyer & Dimick, cited in Anderson et al)
- UHN’s project was driven by the goal of “improving patient safety” (though realizing that other short- and long-term benefit would also be expected)
- in addition to reducing transcription errors and improving order-to-administration turnaround time, expected alerts –> improved quality of care through identifying, at the time an order was entered:
- potential drug-drug interactions
- potential drug-lab interactions
- potential drug-allergy interactions
- also expected to see “inefficiencies from updated order sets and better compliance with drug formulary”
- UHN didn’t collect baseline data for errors – they felt chart auidts would be too costly and take too much time; instead, they extrapolated from the Baker et al (2004) data
The Benefits of MOE/MAR Implementation: A Quantitative Approach
- UHN’s study objective: “to measure the impact of MOE/MAR on patient safety, clinical workflow and the quality of patient care.”
- used study data to “measure implementation progress” and use that to make any necessary adjustments
- if successful, findings can help affect attitudes and perceptions toward future implementation
- shared data collected weekly (or more often) “so that each group could see what improvements there were on their specific units”
- their approach
- steering committee to “determine objectives and research questions”
- select key metrics & indicators
- identify data sources (e.g,. chart audits, time-motion studies, electronic reporting)
- collect/analyze data
- communicate results on a regular basis
- indicators nee to be “meaningful and applicable for clinical and hospital executive decision making”
- collected data for a 2 year period for 8 clinical service clusters (medical and surgical patient care areas)
- some lessons learned:
- good communications can help clinicians understand long-term benefits (to get them through the short-term pain)
- it’s important to manage expectations – e.g., MOE/MAR won’t reduce ordering time, but will streamline process overall; it won’t be perfect
- make sure indicators are useful, keep methods as simple as you can/minimize HR requirements for data collection
- think about data collection issues up front (will system extract the data you need? do you have the skilled personnel to extract and analyze it?)
- get executive commitment and sponsorship
- think sustainment – the evaluation can become “a framework for how people do their work on a daily basis” – continue to use results to inform planning
