How I Learned to Stop Worrying and Love the Data

For more than a year, the service scores at the Moreno Valley Optometry department zigged and zagged in no discernable pattern. Asked whether receptionists were helpful and courteous, 100 percent of patients answering the Ambulatory Service Questionnaire gave the highest score one month.

But two months later, only 78 percent of respondents were that enthusiastic. Two months after that, scores were back up into the 90s. The huge swings were discovered in May 2011 by Stephanie Valencia, the department’s new manager, who excavated two years’ worth of data.

“We had never looked at it before,” she says. “There was no trend. The scores were inconsistent.” Worse, says Valencia, the feedback from the most recent months was headed “on a downhill streak.”

与劳工组织的共同负责人吉娜·希特(Gina Hitt)合作，瓦伦西亚和以单位为基础的团队收集了信息，并设定了基线。吉娜是一名眼镜师，也是Teamsters Local 166的成员。在九月份的两天里，医疗助理询问所有的病人他们是否认为接待员有帮助，然后统计结果。

The team used these to measure the effectiveness of a rapid string of small tests of change. These included adding a smile, positive tone of voice and eye contact on successive days. Each of those days, Hitt and her colleagues asked patients whether their receptionist was courteous and helpful. With each successive effort, the chorus of “yes” got louder and more effusive.

The act of simply examining the service scores seemed to set the team on an upward trajectory: The April 2011 score of 79.55 percent jumped to 89.09 percent in September and then 92.73 percent in October.

瓦伦西亚说:“看到人们如此投入，真是太棒了。”“每个人都是同步的。”

So, that’s a happily-ever-after story, right? Once upon a time, there was an optometry team in Southern California that never looked at its service scores. Suddenly, team members learned their scores were inconsistent and heading in the wrong direction. They focused on key data and tried out small tests of change. Their new practices boosted the score. Everyone lived happily ever after.

Kaiser Permanente的每一个UBT每天都在发生这样的事情。

Right?

Maybe not.

一些UBTs擅长利用数据来指导他们改善绩效的尝试，无论是提高服务分数、减少感染、创造更安全的工作场所还是提高出勤率。但对其他人来说，对数据和数字的恐惧和焦虑是团队通向高绩效道路上的一个重大障碍。

为了符合在绩效之路上的第4级团队的资格，团队必须收集自己的数据，并审查这些数据，看看更改是否有助于提高绩效。为了提升到第5级，团队必须使用带注释的运行图表来测量他们的进度。

But what if you break out in a cold sweat and experience shortness of breath at the sight of anything vaguely resembling math or numbers? Do you simply resign yourself (and your team) to being roadkill on the Path to Performance?

No. Read on.

1) Realize you are plenty smart enough.

Kaiser Permanente, like all large health care organizations, collects and stores vast amounts of data in a variety of complex databases and websites. It employs people with a huge variation in their knowledge of and comfort with data. Just because you’re not at ease with numbers now doesn’t mean you never can be.

就连马萨诸塞州独立非营利机构医疗保健改善研究所(Institute for Healthcare improvement)绩效改善部门的执行董事鲍勃·劳埃德(Bob Lloyd)也开玩笑地把统计数据称为“虐待狂”。

Luckily, the data you will need to turbocharge your team’s efforts to improve performance is probably a lot less complex than you fear.

南加州Kaiser Permanente公司的绩效提升导师迈克尔•默滕斯(Michael Mertens)表示:“这不是真正的公式、统计数据和计算的‘数学’。”“主要是关于前后、加减法。”

2) Whether you acknowledge it or not, you collect data every day.

莫雷诺山谷的眼镜师希特说:“我在变化测试中的角色是征求患者的反馈。”她不需要电脑程序或电子表格。一张纸和一支铅笔就可以了。

“We are all data collectors,” proclaims Stacy Dietz, the UBT consultant for regional operations in Southern California. “And every day, we alter our behavior based on data.” For instance, we ask, “What is the temperature outside?” Then we decide whether to wear a wool turtleneck or tank top. We ask, “What is the length of my commute?” Then we decide whether it makes more sense to drive or take the train.

If you can collect and analyze data to determine your wardrobe, you can also do it to improve the performance of your team.

3) Before diving into the numbers, focus on the “why.”

As the new Kaiser Permanente ads challenge viewers, “Find your motivation.” For unit-based teams, the Value Compass offers a handy cheat sheet on motivation: The patient is at the center. Every data point on every chart represents the impact—positive or negative—that a Kaiser Permanente team had on a patient.

IHI’s Bob Lloyd explains there are three distinct reasons in health care for collecting and examining data:

• 比如KP最近的一项研究发现，60多岁时骨折的女性在一年内死亡的可能性是没有骨折的女性的五倍。
• 在判断方面，一个类别将包括联邦政府最近对医疗保险计划的质量和服务排名(有几个KP计划得到了五星)。该类别还包括决定医疗中心或部门是否获得其绩效分享计划(PSP)奖金的分数。
• 改进的。

This last is the reason UBTs should be collecting and examining data.

“在质量改进工作中，测量的目的是为了学习，而不是判断，”劳埃德说。

数据可以回答诸如“我们现在做得怎么样?”“随着时间的推移，我们会变得更好吗?”还是变得更糟?”“我们对改变的小测试有效果吗?还是不是?”在缺乏数据的情况下，我们倾向于依靠猜测、直觉、趣闻——无论是否合理，我们都倾向于责怪或赞扬特定的个人。

“You need data. Otherwise, you don’t have any solid information,” Hitt says. “You just have word of mouth.”

4) Only gather the data you actually need.

The holy grail of data for UBTs is the run chart. Don’t let the name throw you. It’s simply a chart that tracks some number (say, a service score, or number of last-minute sick calls) over time (day, week, month, quarter).

“The most crucial question to ask is, ‘What are the few, vital pieces of information that are important?’ ” says Dennis Benton, executive director of the Panorama City Medical Center in Southern California. Any graph or data set that requires its intended audience to get special training to read is probably too complex for the task at hand, he says.

本顿说:“你可以在UBT会议上进行快速、及时的培训。“我们在领导轮会中这样做。我指着图表，谈论它们。”

Run charts make it clear at a glance how your team's tests of change are working.Use this toolto walk through how to make one.

4 1/2) But, get the data often enough to support your improvement efforts.

For most teams’ small tests of change, data that can be collected daily, weekly or—at most—monthly will be most useful. Waiting for quarterly reports is generally not going to cut it. The Moreno Valley Optometry department did not wait for the Ambulatory Service Questionnaire results—which are posted monthly—to come in. It’s called the Rapid Improvement Model, folks. Not the Slow-as-Molasses Improvement Model.

Bottom line: The data should be useful for the team and be determined by the team.

5) Think art class, not math class.

“I hate numbers,” admits Jenny Yang, a receptionist at the Moreno Valley Optometry department and a member of the UBT’s representative group. When the notion of using service scores to guide improvement first came up, Yang says she told her teammates, “I’m not going to do it. Make someone else do it.”

To help others like Yang, Benton says, when it comes to data, “Make a picture out of it. I am a big believer in graphs. With a graph, you can say, ‘We dipped here. What is the reason? What can we do about it?’ You can look at a trend relative to the goal.”

“Graphs are visual,” Valencia adds. UBT members have a variety of learning styles and preferences: “Everyone learns differently.”

And think in terms of moving video, not still photographs that capture single moments in time. IHI’s Lloyd asks, would nurses measure an ICU patient’s vital signs only when the patient arrived and when she left the unit? Or would they monitor vitals constantly via a telemetry machine? The second option is better, so caregivers can intervene in real time to help the patient’s recovery.

6) You didn’t like art class?How about creative writing?

数字可以说明问题。《幻灯片:学》(Slide:ology)和《共鸣》(Resonate)的作者南希·杜阿尔特(Nancy Duarte)说:“数据中有叙述。”这两本畅销书讲述了如何进行引人注目、令人难忘的演讲。“是什么让数字上下波动?”这些数字有多大?这些数字与其他信息相比如何?”

Yang agrees. Graphs with data “give you key points, high points and low points and trends,” she says. As a member of the representative UBT, Yang—a member of Steelworkers Local 7600—sees herself as a storyteller: “My audience is the UBT. The graphs help UBT members make sense of everything.”

Hey, if you liked math class, more power to you. “I love math,” says Hitt. “I am a number cruncher. But for me, charts and graphs? Not so much.”

7) It’s OK to ask for help.

So that graph you pored over in your UBT meeting is still making you break out into a cold sweat?

“It’s OK to find a safe place to say, ‘I don’t get this,’ ” says UBT consultant Stacy Dietz. That might not be in a big group, but it could be one on one with a trusted peer.

Mertens, the Southern California performance improvement mentor, says the best way to learn to use data is to try it out. At the request of Susie Bulf, a UBT consultant, Mertens led a training for UBT co-leads in Fontana on how to create a run chart. He led an in-class exercise using sample data—and then another exercise where each team used its own data.

默滕斯说:“第一次做就能克服焦虑。”

Each KP region boasts a roster of experienced performance improvement mentors. In addition, most UBT consultants have had some training in performance improvement strategies.