The Mihalik Blog

Quality Improvement 101: Affinity Diagram

Creating an Affinity Diagram is a fun way to organize facts or ideas.  It works especially well when there seems to be chaos.  It can be useful in generating broad themes and can definitely catalyze creativity and breakthrough thinking. 

 

An affinity diagram, groups related ideas together.  Unlike an Ishikawa (fishbone) diagram, the ideas are not necessarily causally related.  Whenever I see an Ishikawa diagram that doesn’t extend to more than two branches (a big branch off of the main spine, and a smaller branch off of that one), it usually seems to me that an affinity diagram would have been a better solution—and a lot more fun to generate.

 

Gather together a team of people who are familiar with the topic to be addressed.  They should have the requisite knowledge and experience and be accustomed to working together but avoid “the same old crew.”  We’re aiming for creativity here so some new blood will be helpful.

 

Broadly state the issue to be addressed using neutral language.  For example, use the word issue rather than problem.  Use brainstorming  to generate ideas.  Be sure that when you complete the clarification step, each statement has a verb and a noun if possible. 

 

After the ideas have been finalized, write each one down on a large “Post-It” note.  Using a large section of blank wall, stick up the notes in random order.  In silence (yes, really) the team should begin to move the notes into related groupings.  The ground rules, in addition to no discussion, are react more than contemplate, move the card if you disagree, and try to avoid “known” groupings.

 

There will typically be 6 to 10 groupings when you’re finished.  For each grouping either find a card in the group that would work well as a header or make one.  Header cards should be three to five words long and avoid jargon or comfortable terms.  An ideal header would be able to stand alone and convey the essence of the group. 

 

To display the affinity diagram, draw boundaries around each group and its header.  Place related groupings near each other and connect with a line.  If appropriate, you can create a super-group of two or more individual groups and develop a super-header.

Quality Improvement 101: Benchmarking

I want to talk about a pet peeve of mine—the misuse of the words benchmark and benchmarking. 

 

It seems that folks often use the term benchmark and goal interchangeably.  They’re very different. 

 

A benchmark is the best of the best of actual performance.  It is a level of measured performance that some organization has actually achieved.  A benchmark cannot be “set” the way a goal can.  You get to choose your goal.  A benchmark is something that “is.”  In this context, one might define benchmarking as the process of measuring an organization’s performance on a key customer requirement against the best in the industry, or the best in any industry. 

 

The last little twist (“the best in any industry”) is often a characteristic of benchmarking efforts.  You may need to go out of your industry to find a benchmark.  For example, it is unlikely that the benchmark for customer service would be found in a health care organization.  You might want to consider Disney or Nieman Marcus in that regard. 

 

Usually, I consider benchmarking to relate to the process of measuring an organization’s performance against the best known performance.  Another definition of benchmarking, though, is the search for industry best practices that will lead to exceptional performance through the implementation of these practices.  This approach to benchmarking focuses not on the level of performance itself, but what the actual practices are that lead to exceptional performance. 

 

Notice that this second definition of benchmarking does not actually include consideration of the measurement of performance.  It speaks more to evaluating the processes of another organization and adopting those that are considered industry best.  To satisfy NCQA requirements for either a benchmark or a goal, you need a number so this approach to benchmarking alone will not be adequate without the adoption of a quantitative goal.

 

Another way to define benchmarking is the practice of being humble enough to admit that someone else is better at something than you are and being wise enough to try to surpass them at it! 

 

We’re always here to help you achieve your performance objectives. 

Quality Improvement 101: Selection Grid

A Selection Grid is another way to establish priorities among an array of options.  It is a more structured process than either multi-voting  or nominal group technique.  It is used when there are clear criteria for making a decision.

 

One of the major conceptual differences between using a selection grid and using multi-voting or nominal group technique is the emphasis on conscious decision-making.  Explicitly articulated criteria are not used in multi-voting or nominal group technique. 

 

There are many selection grid methodologies.  Some can get quite complex but have the advantage of being able to establish priorities among a large number of options.  In health care, I generally use a consensus-based method, which can be time-consuming and difficult to use for more than about 8-10 options. 

 

In order to use a selection grid, there need to be decision-making criteria.  If you do not already have criteria (some organizations already have criteria to be used when making certain types of decisions even if the criteria are not used in the context of a selection grid) you may want to consider multi-voting or nominal group technique unless you have a good reason for wanting the decision-making criteria to be explicitly stated.

 

As with the other techniques, pull together a representative group of participants.  Clearly state the question and identify the options from which the selection is to be made. 

 

Construct a grid that lists the decision-making criteria down the left side and the options to be decided on across the top. 

 

Using a scale of 0 to 9, rate each option on each criterion by group consensus.  Alternative methods allow each individual to do his/her own rating and to average the results to arrive at the rating for each cell in the table.  It is also possible to weight the criteria so that more important criteria carry more decision-making weight. 

 

When all the ratings are done, add up the score for each option and put the scores in a row across the bottom of the table.  Discuss the ranking.  Does it make sense?  Remember that small differences in total score may not be significant but large differences almost certainly are. 

Quality Improvement 101: Nominal Group Technique

Like, multi-voting, Nominal Group Technique is a method of paring down a broad list of ideas and establishing priorities. I use nominal group technique more than multi-voting. I suggest you try both a few times and decide which you prefer.

Gather together a representative cross-section of individuals, write the ideas on flip chart pages, and letter each idea just as you would for multi-voting. Most likely. the ideas will come from one or more brainstormingsessions.

Now, ask the participants to write down the letters of the five ideas that they think are most important on an index card or piece of paper. By now, if you’ve read the last few entries in this series, you know that this process should happen in silence.

The facilitator then asks each participant to identify the most important idea from his/her five ideas and to put the number “5” next to it. Allowing time for each step so that no one is rushed, the facilitator asks the participants to put the number “1” next to the least important of the remaining four items; then a “4” next to the most important of the remaining three items; then a “2” next to the least important of the remaining two items, then a “3” next to the lone remaining item.

The cards are passed into the scribe who, usually with an assistant, transcribes the votes to the flip chart adjacent to the appropriate idea. Before discussing the ranking, the group should decide if the results seem to have adequately pared down the list to a group of high-priority items. If not, the process can be repeated after eliminating all items that got no votes and probably those that got only one vote.

A big difference between the analysis of the results of nominal group technique and the results of multi-voting is that the group considers the total point count for each item (adding up the values of each individual vote) as well as the number of votes that each item got.

Consider the difference between an item that got five votes of “1” and an item that got one vote of “5.” In the former instance, five different people put the item in their top five but in the latter, only one person felt the idea made the cut into the top five. Don’t dismiss the single vote of “5” though. Consider, for example, whether the individual who cast the vote has a unique perspective on the process that prompted her/him to rate that idea as the most important.

It can be particularly difficult to facilitate sessions like this since the process is unfamiliar to most individuals. There will be a tendency to want to switch into “committee mode” and begin discussing ideas. I suggest you rigidly adhere to the rules to get a sense of what the technique can accomplish.

Quality Improvement 101: Measures

In the world of NCQA accreditation, HEDIS measures get lots of attention.  Sometimes it is easy to loose focus on the fact that there are many measurement requirements that are not included in HEDIS.  Among others, measurement is required for Availability of Practitioners; Accessibility of Services; adherence to Clinical Practice Guidelines; Continuity and Coordination between Medical and Behavioral Healthcare; Clinical and Service Improvements (for MBHOs); and for Health Plans Satisfaction with Case Management, Effectiveness of Case Management, Satisfaction with Disease Management and Effectiveness of Disease Management. 

 

And that’s only a short list!!

 

Given the importance of measurement, I thought it would be useful to go over a few key aspects of measures. 

 

In general, measures are better if they are direct measures of actual performance that are proximate to the event of interest rather than indirect or proxy or distal measures of performance.  Although I see it stated frequently, readmission rate is not a measure of poor discharge planning—it is an indirect, distal measure that is a poor proxy for the topic of interest. 

 

It is always better to measure performance in actual units (such as seconds to answer the telephone) rather than measuring whether the standard was met (such as telephone answer in 30 seconds or less).  You can always convert the former into the latter, but not the reverse.  If you were to bid on a new contract that included penalties for a speed of answer of greater than 20 seconds, and you only knew if you could answer the telephone in 30 seconds, you would not be able to determine the likelihood of incurring a penalty. 

 

Consider whether your measure is evidence-based.  Has it been validated?  Is it a published measure or a home-grown one? 

 

Good measures have known properties.

§         Validity:  Does it measure what it proposes to measure?

§         Accuracy vs. Precision:  Accuracy refers to how well the measure conforms to an absolute standard of measurement whereas precision refers to the ability to detect change from one measurement to another.  For example, my bathroom scale may not give me an accurate weight in absolute terms but it may do perfectly well at letting me know how much weight I’ve gained or lost while on vacation.

§         Reliability:  Does the measure produce consistent results.  For example, if I measure the same event a second time will I get the same result? 

 

If you need help constructing measures or evaluating performance we can help. 

 

Quality Improvement 101: Multi-voting

Multi-voting is a technique to pare down a broad list of ideas and establish priorities.  Typically, multi-voting follows brainstorming or some other idea generating process. 

 

As with brainstorming, you want to gather together a representative cross-section of individuals so that the results are well balanced.

 

Write the ideas for the group to consider on flip chart pages.  If you just finished a round of brainstorming, that’s where the ideas will come from.  I suggest lettering each option rather than numbering so as to reduce confusion with the numeric results of the multi-voting. 

 

Give each participant a specific number of votes to cast.  I suggest using five votes to start but you can change this number as you gain experience in using the technique.

 

In silence (yes, really) and using an index card or piece of paper, each participant writes down the letter of an idea for which he/she wants to vote and the number of votes he/she is casting for the idea.  I suggest using hash marks rather than numbers for the vote count.  The cards are passed into the scribe who, usually with an assistant, verifies that each card contains the right number of votes and then transcribes the hash marks to the flip chart adjacent to the appropriate idea. 

 

 

The scribe counts up the number of hash marks for each idea and writes the number down next to the idea.  If necessary, the process can be repeated after eliminating all items that got no votes and probably those that got only one vote. 

 

When the multi-voting is over and the results tallied, the ranking is discussed and analyzed.  One question to consider is whether the results seem reasonable.  In my experience, if the group is well-selected, the results will be credible. 

 

If you’ve followed this process through from brainstorming to arriving at the short list of high-priority items by multi-voting, you’ll notice that the first time for real group discussion is only after the prioritization is complete.  Not only does this make the process faster, it really will generate many more ideas than the typical committee discussion paradigm. 

 

A few other thoughts:

§         One of the ground rules is to decide how the maximum number of votes someone can cast for a single idea.  I usually suggest that each individual must cast votes for at least two ideas, preventing them from casting all their votes for only one item.

§         It is possible to allow individuals to walk up to the flip chart pages and cast their votes directly by making hash marks next to the items they’re voting for but you’d be surprised how often folks forget how many votes they’ve cast before they’re done.

 

 

 

 

 

 

 

Quality Improvement 101: Brainstorming

We use the word brainstorming all the time but how many of us really view brainstorming as a powerful technique to encourage unrestrained collective thinking?  I’ll bet that most of us just think brainstorming means that we think about something a bit, maybe chat about it in a group, and come up with some ideas. 

 

Not.

 

While there are a number of ways of doing brainstorming, I’ve found the following technique to work well in a wide range of circumstances. 

 

Gather together a group of about 8-12 people who are knowledgeable about the topic.  The goal is to get a wide breadth of representation from across the organization and even from external stakeholder groups if appropriate.  The scope and creativity of the ideas generated will be related to the size and composition of the group. 

 

Define the topic clearly and write it down for all to see.  I find that using a flip chart is good for this purpose.  Keep the topic concise but clear, something like:  “Ideas about how to improve telephone responsiveness.”

 

In silence (yes, really) have everyone write down every idea that he/she comes up with that relates to the topic as defined.  This usually takes 5 to 10 minutes.  Individuals who finish earlier than others should sit in silence (yes, really) so as not to disturb those who are still working.

 

Appoint a scribe.  Rotate through the group, in a defined order, giving each person the chance to state one of his/her ideas.  This is not a time for the individual to justify or support the idea, just to get the idea up on the flip chart.  The scribe should write down each and every word (yes, really) uttered by the other person.  The job of the scribe is to write, not to edit, comment, editorialize, or anything else.  On the other hand, the person who is stating idea has to state it succinctly so that it can be written down.  The facilitator or the scribe should ask the individual to restate the idea if it is too long to write easily. 

 

Rotate through the group as many rounds as needed for everyone to state each of his/her ideas.  There will be a tendency for individuals to indicate that many of their ideas are already up on the flip chart.  Insist gently (yes, really) that they still put their ideas up there.  If an individual runs out of ideas, he/she just drops out of subsequent rounds, which continue until all ideas of all participants are up on the flip chart. 

 

Once all ideas are transcribed, have the scribe or facilitator read each idea and verify that it is clear to the group.  This is not the time to discuss the merits of the idea (or anything else for that matter).  Just be sure that everyone understands what the idea means.  If not, ask the person who came up with the idea to edit it so that it is clear. 

 

After all ideas have been clarified, the group can combine ideas that are essentially identical.  Be careful, there will be a tendency to lump similar ideas together.  Don’t do this.  You only want to combine duplicate ideas into a single idea and only if you have the permission of the individuals who came up with the original ideas. 

 

For some techniques to establish priorities among the ideas read my upcoming entries on Multi-voting and Nominal Group Technique.

 

A few final thoughts:

§        Once the process has begun, don’t allow latecomers to join the group.  It is likely to disturb the process.

§         Create an atmosphere where all ideas are welcome and criticism is absent. 

§         Gently but firmly redirect anyone who tries to dominate or derail the process.   

Quality Improvement 101

In a few blogs I’ve lamented the fact that serious QI doesn’t seem to be happening as much as it did a decade ago.  Fresh from teaching two NCQA educational programs, one on Data Analysis for Quality and Process Improvement and the other on Behavioral Health HEDIS Improvement, I thought I would write a series of blogs on the tools and techniques of QI that I find most useful. 

 

Along the way, I’ll include some entries on other aspects of quality improvement and performance measurement that I think are worth mentioning.

 

The QI tools and techniques that I plan on addressing in a “how-to” fashion include:

§         Brainstorming

§         Multi-voting

§         Nominal Group Technique

§         Selection Grid

§         Affinity Diagram

 

My hope is that you will use some of these techniques to catalyze your improvement efforts.  Far too often we just sit in committee meetings and have the same discussions over and over again.  If we really want to jump-start our improvement efforts (say for example, making those HEDIS measures pop) we need to do something different from what we’ve done thus far.  If we keep doing the same old thing, we’re going to keep getting the same old results. 

 

I encourage you to try each of these techniques.  The results just may surprise you.

New Year’s Resolution: Planning for Improvement

Our office reopened on January 2^nd after a year-end break. 

Late December into early January is a period when not much new happens in the areas of accreditation and regulatory compliance.  That makes it a very good time to plan for the coming year.  The quality improvement calendar is a critical part of the planning process—and one that I find is often not used to its full potential. 

The problem that I typically see is a timing issue.  QI program evaluations and revised program descriptions are typically not completed until late first quarter or early second quarter.  Many health care organizations include the annual QI calendar as part of the revised program description.  Consequently, the QI committee has not explicitly planned activities for the first three or four months of the year. 

As a New Year’s Resolution, I suggest you do away with treating the QI calendar as a once-a-year task.  Quality Improvement is a dynamic process and planned activities should be modified during the year based on new information. 

Set up your QI calendar in a table format.  Each activity should occupy one row.  Set up columns for the information you want to include, such as the name of the indicator, the responsible party, the goal, etc.  End your table with a series of twelve columns, one for each month.  In each of these cells, describe what should be happening during that month for the specific activity.  Leave the cell blank if nothing should be happening.  Developing a set of abbreviations will make it easier to maintain the calendar; things like M for measure, R for report to committee, etc.

On a regular basis, as you are preparing the minutes of the most recent QI committee meeting, add enough new columns to the calendar to project a full twelve months into the future.  In the new cells, describe what should be happening during each newly added month, if you know.  For example, if a particular indicator is reported on a quarterly basis, you can project out when the next reports will be due. 

Every three to six months, make the calendar update a specific agenda item for the QI committee.  With the committee’s input, revise the calendar to customize it to the reality of what is happening. That quarterly report you added might need to be revised if there’s some sort of a delay, for example.  If you schedule one of these calendar updates for the last QI committee meetings of the year, you’ll be starting the new year with the most up to date plan possible. 

Although it is easier to word process in a word processing program, using a spreadsheet program for the calendar will enable you to print just the columns that you want to print.  So, while your spreadsheet might have two years’ worth of planned activities, you can choose to print just the next twelve, without loosing the history.

Making Quality Happen

I just got back from a brief jaunt to the East Coast where I gave a presentation on HEDIS and Quality Improvement and then facilitated a small group session to identify interventions to improve performance.  All in all, this is a fairly typical role for me.

What struck me for some reason this time around, was how much more often I used to facilitate quality improvement processes than I do now.  I’m not sure what caused the change but I’m more convinced than ever that quality improvement efforts at health plans and health care facilities would improve significantly with good facilitation.  Honestly, I think that the health care field isn’t paying enough attention to good technique when it comes to the tools and techniques of QI. 

As part of my talk, I described brainstorming techniques, much as I did in a previous blog.  Despite the “just in time” training, and knowing that I was sitting in the back watching, the groups were not able to implement what they had just been taught.  Brainstorming sounds simple but learning to do it well is like learning many other skills.  It takes practice, modeling, and mentoring. 

The group members quickly fell into the all-too-common pattern of having long, discussions about their thoughts of what would improve performance.  When I was asked to help facilitate, I took a moment to refocus the group on the task and then went back to the book.  The book on structured brainstorming, that is. 

I had everyone take 5 minutes to write down his/her ideas for possible interventions to improve performance.  We spent the next 25 minutes transcribing everyone’s ideas onto the flip chart.  On a number of occasions, had I not “facilitated” the process, the group would have fallen back into the old habit of discussing ideas. 

In the course of 30 minutes the group had generated over two dozen credible interventions to improve outcomes for a specific high-risk patient group.  In the thirty minutes before I began facilitating, they hadn’t generated any. 

Clearly the ideas were there—they just needed to be set free.  Good facilitation was the key. 

Although it might sound self-serving, I truly encourage you to think about involving The Mihalik Group in deployment of your quality improvement program.  When you need to understand the causes of key outcomes or develop interventions to improve performance (like maybe your HEDIS rates), our facilitation skills may be just what you need. 

Give me a call or send me an email and we can discuss how The Mihalik Group can help.