Key Metric

Loading...

En provenance du cours de Johns Hopkins University

Taking Safety and Quality Improvement Work to the Next Level (Patient Safety VII)

13 notes

Johns Hopkins University

Taking Safety and Quality Improvement Work to the Next Level (Patient Safety VII)

13 notes

Cours 7 sur 7 dans Specialization Patient Safety

In this culminating course in the Patient Safety and Quality Improvement Specialization, you will apply the skills you have acquired across the previous six courses to address a realistic patient safety issue confronting Mercy Grace, a 500-bed urban hospital that is part of a larger hospital system. Based on the scenario provided, you will assess the situation and work through the problem using a variety of tools and strategies. You will have the opportunity to identify defects, root causes, and potential mitigation strategies; you will create a project implementation plan for addressing the issue in the form of an A3; you will identify risks of project failure and design a change management plan; you will identify means of converting the project from local to system-wide; and you will identify quality and safety measurements that will be used in evaluating the success of the project’s implementation.

À partir de la leçon

Problem-solving and Proposing Interventions

In this module, you will build upon what you learned in the previous module by continuing your investigation of patient falls and clarifying your project goals. You will review content from previous courses in the specialization that discussed problem identification, scoping your project, building an A3, and specifying performance measures. You will then initiate a first draft of an A3, and complete a first draft of a 5-Whys Table.

Module 2 Introduction1:02

So, What Is Your Problem?10:33

Building Your A35:40

Specifying Performance Measure and Common Data Sources for Measurement10:45

Rencontrer les enseignants

David Schwartz
Instructional Design Technology Manager
Armstrong Institute for Patient Safety and Quality

Hello. We're now going to talk about the Key Metric in

terms of project selection and why the Key Metric is so critical.

When you think of a Key Metric,

it's used to determine baseline and ongoing performance.

So, we measure it several times along the way.

First, we want to anticipate what we think that Key Metric is going to be.

Another term for Key Metric is what we call,

KPI or Key Performance Indicator.

You get one, there's only one Key Performance Indicator.

You have to understand what that is in

terms of stating the extent of the current problem,

and also a progression towards realizing that goal.

It has to be clearly defined.

So it's understandable.

Examples could be things like, defect rate.

So, perhaps it's your falls rate,

perhaps it's a compliance issue that we're going after,

perhaps it's a time issue in the way of throughput.

These are things we'd want to understand and be able to articulate.

So, we have the problem statement,

now we have to have a way of measuring the problem and that's what this Key Metric does.

Support metrics, you hear support term,

support metrics bandied about a process metrics.

They're perfectly acceptable, in fact I encourage them.

But again, there should be only one Key Metric to

gauge baseline and improvement following up.

So, let's give an example of a Key Metric.

It could be the percentage of patients left without being seen.

I'd even go a little bit further in explaining that Key Metric.

Any time I see a percentage,

I'd like to know what you think the numerator and denominator of that would be.

So, perhaps it's the number of patients left,

divided by the number of patients that were registered or perhaps triaged.

The support metric might be something a little different.

The support metric helps you navigate the waters to

understand some of the details surrounding what's driving that Key Metric.

So, maybe, we think the element of time

is one of the factors in terms of why patients are leaving before they're seen.

So, we might want to track the median number of minutes from triage to treatment.

Some considerations in terms of Key Metrics.

Is the data you need already available?

This is a huge barrier that we see all the time where people articulate,

yes I think the key measure should be this,

and then we find out that there's no potential way to obtain that data.

If that data is available,

who has access to it?

Do you have permission to access it?

Will they help you access it?

Will it need to be manually collected?

All considerations you need to think through.

If it does, are you going to do it all?

Are you going to engage with the help of others?

That makes a big difference.

How much data will you need?

There are a lot of ways you can calculate sample size and things like that.

But just to keep it simple,

the two things that initially I look at,

are what's the level of variation in your process?

And is there any seasonality?

So for example, a while back,

we had had a problem with blood wastage.

And we realized that our volumes of blood wasted and blimes of blood issued,

varied on the seasonality because where I live in the mid-Atlantic,

people in the summer tend to get and be out more and they engage in

activities where they hurt themselves and have issues and they require more blood.

So, in the winter months,

our blood usage drops,

and the summer months,

it is much more dire and they need more blood.

How much data formatting is required?

Are you an Excel whiz?

Are you good with spreadsheets?

Well, these are the types of things you want to look at.

Do you need help with that?

Or do you need training in that?

So, you want to keep it simple,

that's the underlying way to do this.

But again, you might need some help in terms of taking that rough data,

and carving it down to what you need.

Next, let's talk about operational definition,

or how we will obtain the data.

It's literally the who, what, where, how,

when, of how we're going to obtain that metric.

It explains how to obtain the value of that.

It describes exactly how you're going to obtain the measurement from the data.

We want to make sure it's meaningful to your process.

We want to make sure it's as precise,

and procedural as it can be.

So for example, if you have three people out there collecting the data,

you need to ensure they're doing all the same way.

Let's give an example of an operational definition.

So, in the surgery clinic this could be,

"Lead time is measured in minutes from the time the patients signs

the log through the patient departure time as noted by who?

The Clinical Associate on the patient chart."

So, the lead time is somewhat the what?

The how it's how is measured in minutes.

The who, the patient signs a log?

When and where, from the page for action time.

This is noted by the Clinical Associate on the chart.

So, it's the who, what, where, how, when,

that's what we want to try to obtain.

Manual data collection.

So as I said earlier,

sometimes that data aren't available.

So, you have to go or you have to have someone go collect that data.

You have to understand,

who is going to be engaged in this activity and get

air coverage for them to collect data because they're doing other jobs at the same time.

You have to have your operational definition to create

a data collection tool to build a know how this data will be collected.

You might have to train folks.

And a lot of times what happens is if you have

more than one individual collecting that data,

what I recommend is you go out with them and

each individual person is witnessing and they're capturing

their own data and then you can compare to ensure that

you have that reproducibility going on there.

So, you want to basically want to have reproducibility

and be able to articulate how we did this and why.

Create spreadsheets.

So now that we have this data,

you have to have a way to collaborate together if you will.

You have to have a way to summarize it and make it meaningful.

So, you might need a spreadsheet to do that,

you might need to create a chart,

or something along those lines.

Key metric do's and don'ts.

Do, have others "weigh in".

You might have noticed this pattern already that,

every single do starts with have others "weigh in" on this.

Have your stakeholders "weigh in",

and see if it makes sense.

You have to ensure that the data needed is attainable.

If you can't get it it's not going to help you.

You have to ensure that the process relates to both the problem and goal.

You only have one problem,

one key element of a problem,

and one goal that relates to that problem.

Include both numerator/denominator to explain a percentage.

And identify that operational definition

we just talked about and data collection mechanism.

Some of the things I try to avoid are,

using action words like improve, reduce.

The metric is simply a metric, it's a measure.

So, there shouldn't be any verbs in there.

Choose something you can't measure or measure with accuracy.

So, a while back we had an issue with medications being unlabeled.

And we had a heck of a time understanding how we

were going to capture the rate of unlabels.

So, it's one of those things we had to put a lot of time,

energy, and effort into.

And we really weren't overly successful and understand the data.

Casting it in stone.

Don't think that once you establish that key metric,

it will be the key metric.

It's going to morph like the problem statement and go well over time.

Express in dollars, unless it's all about the money.

Focus on the clinical benefit and

possibly list the dollars as an added benefit in the end.

Explorer notre catalogue

Rejoignez-nous gratuitement et obtenez des recommendations, des mises à jour et des offres personnalisées.

Coursera propose un accès universel à la meilleure formation au monde, en partenariat avec des universités et des organisations du plus haut niveau, pour proposer des cours en ligne.

© 2019 Coursera Inc. Tous droits réservés.

Télécharger dans l'App Store

Disponible sur Google Play