How to win, hold support for measurement initiatives
How to win, hold support for measurement initiatives
Skill building and common sense are critical
By any name — data collection, survey, study- — measurement has a bad reputation. People know too well that it frequently does more good for somebody else than for them. Often, says Charles Kilo, MD, senior fellow with the Institute for Healthcare Improvement in Boston, and chief medical officer for MyHealthBan in Portland, OR, data collection precedes staff cuts or increased workloads.
Judy Homa-Lowry, RN, MS, CPHQ, president of Homa-Lowry Healthcare Consulting in Canton, MI, observes that project leaders sometimes set employees up for failure in data collection efforts by assuming that they know the fundamentals. Briefing the project team on data collection methods and goals can secure their cooperation and save time by capturing useable information from the beginning.
Train for success
Sometimes health care workers need to revise or augment their academic training to competently participate in measurement initiatives. "By and large, we underuse their abilities," observes Kilo.
Not everyone understands what a random sample is, observes Homa-Lowry. The usefulness of before-and-after analyses taught in medical school is limited, adds Kilo. Many physicians need to learn how to plot events by the hour, day, week, or month to track variance over time as required in ambulatory care or surgical practice.
Kilo explains that for process improvement, either computerized or paper and pencil data collection can be valid. "With a little training," he says, "they can do valid data collection." It’s good practice to train each project team on data collection and frame it either as review or skill building.
Training should also cover:
• Explicit definitions of data elements. Either home-grown definitions or those adapted from other sources are acceptable as long as everybody understands and uses the same ones. For example, the length of an office visit could extend from the time a patient arrives until she leaves; or it might extend from the time the patient walks into the exam room until he leaves the office.
• Exact sampling times, such as week #1, Monday, 1 to 3 p.m.; week #2, Friday, 10 a.m. to 12 p.m.
• Time lines for recording and turning in data.
To build long-term alliances with your stakeholders and project participants, figure out early how to reward them. What value could this data collection have, for instance, to the bedside nurse, staff pharmacist, attending physician, maintenance worker, or receptionist?
If the aim is better pain management of orthopedic patients, involve your team in targeting the care practices that need improvement, collecting baseline information, and designing measures to track the progress. The people closest to the problem could alert project leaders to the point — if leaders demonstrate that they are receptive, observes Homa-Lowry.
Losing sight of the reason for measuring leads to what Kilo calls a "data frenzy." Given latitude to use their common sense, people can define what’s enough as opposed to overkill. For example, with cycle times for a patient care process, sample two events a day. That’s 14 a week or 56 a month. It’s enough to track a rate or a trend, Kilo advises.
For data collection that pertains to strategic plans, such as "reduce the length of stay," drill the goals down to each level and function. For physicians, it could entail writing discharge orders on time; for staff nurses, educating patients on home care practices; and pharmacists, filling prescriptions in timely fashion.
Run a tight input-feedback loop
Present results often and display them prominently, Kilo suggests. This conveys to patients and visitors an organization’s commitment to improvement. "While you might not want to display all your financial data to patients, the patient and staff satisfaction rates and wait times could be posted on a board in the waiting room." he concedes. However, he does encourage posting financial data for employees, such as in the staff lounge.
When your input-to-feedback cycle is short, stakeholders learn to watch for changes. Not only do timely updates reward their adherence to deadlines, but they also sustain interest and keep thoughts of improvement alive, Kilo explains.
Choose clarity over glitz in visuals
Often the best answer to data presentation lies in training one person who has some aptitude and willingness to learn the techniques. This could be a department head, office manager, or administrative assistant. While the computer tools become easier to use all the time, people still need to learn how to construct simple, useful graphics. The most effective learning situation is a class where students bring their data and have somebody teach them the principles of presentation, notes Homa-Lowry, who teaches data interpretation and presentation for the Joint Commission on the Accreditation of Healthcare Organizations in Oakbrook Terrace, IL, and at Madonna University in Livonia, MI.
"I’ve seen presentation of very slick, attractive graphics that capture the audience’s attention, but they don’t give them meaningful information about how it applies to them. We call this infotainment.’ It’s best if people can take a quick look and say, I understand,’" she explains.
Sometimes, this is something as fundamental as noting the time frame to which the data apply. Other fundamentals include clear labels on the vertical and horizontal axes of a graph or conversion of numbers into percentages so users can see changes over time. "If people are not comfortable with their understanding of the information, they’re not going to be comfortable using it," she warns.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.