The Pulse of Progress: Implementing Adaptive Monitoring

Welcome to Episode 23 of the Community Health Management Design podcast series. In today's episode, 'The Pulse of Progress: Implementing Adaptive Monitoring,' we're diving into the art and science of creating responsive health programs that truly serve community needs. Imagine knowing exactly how your program is performing at any given moment, being able to spot trends before they become problems, and having the data you need to make confident decisions about program adjustments. That's the power of adaptive monitoring systems, and today we're breaking down exactly how to build one - no matter your program's size or technical sophistication. From HIPAA-compliant data collection to creating meaningful feedback loops, we'll walk through every step of developing a monitoring system that enhances your program's impact while protecting patient privacy.

Tami Moser [00:00:00]:
Welcome to the Community Health Management Design Podcast. I'm your host, doctor Tami Moser. And today, we're diving deep into the world of adaptive Marna Dering systems, a crucial component of modern health care management that can seem daunting, but doesn't have to be. Whether you're working with sophisticated electronic health records or simple spreadsheets, I'll show you how to create effective monitoring systems that work within your resources and constraints. Before we dive in, let's acknowledge the elephant in the room, data security and HIPAA compliance. These aren't just bureaucratic hurdles. They're essential protections for our patients and our organizations. Throughout today's podcast, we'll explore how to balance the need for real time monitoring with our obligation to protect sensitive information.

Tami Moser [00:00:44]:
First, let's talk about who needs to be at the table when designing your monitoring system. You'll want to connect with your IT department or technology partner. They'll help assess what's technically possible, guide you through security requirements, and help integrate different data systems. You may want to talk to the privacy compliance officer. They're essential for HIPAA compliance review. They can help structure data access protocols and guides acceptable use policies. Department managers provide insights into workflow impact, help identify key metrics, and ensure buy in from their teams. Now this could be an important person to talk to about the other essential metrics that you may wanna design for evaluation from our last podcast.

Tami Moser [00:01:31]:
They can also often include finance personnel. So you may have a finance team or someone in the finance office. They help you with budget considerations, cost benefit analysis of technology, and resource allocation guidance. Now let's explore three levels of implementing real time data tracking from basic to advanced. And let me take one step back first. You may need to stop here and brainstorm who in these different categories you need to talk to and then try to define the person you want to talk to. And if there's any other people you feel that's vital to involve surrounding technology, make a list before we go any further just so you have that as your starting point for discussions. Now let's explore three levels of implementing real time data tracking.

Tami Moser [00:02:21]:
We're gonna start at basic and then move through some advanced implementation. So if we start at level 1 basic implementation implementation, this would mean use of existing tools like Excel or Google Sheets with appropriate security measures. These would be set up for manual data entry at set intervals. You'll have simple dashboards using built in charting tools with Excel or Google Sheets. This is best for small programs or pilot projects. Cost is minimal to none. Most of you are already gonna have easy access to both of these, or at least one or the other will be a part of licensing for your particular organization. And the time to implement is days to weeks depending on your comfort level with Excel and Google Sheets, and how easily you're able to build out the spreadsheets you need.

Tami Moser [00:03:14]:
And, again, this could be someone that you bring on to the team from some of the other areas we discussed who could help you set this up much faster than if you try to do it on your own. Level 2 is intermediate implementation. So this is where you might have dedicated database software or basic health care analytics platforms that you're able to use for this. You can have semi automated data collection and interactive dashboards, which are very helpful. You can also program this for regular automated reporting, and it's best for medium sized programs or growing initiatives. The cost is moderate. Sometimes you may need to invest in additional software, and at other times, you may have software on hand that's already being used that can be adapted for this level of implementation. And it can take weeks to months, again, depending upon what you have and have to purchase and then how smoothly integration can work within your, software packages, if you will.

Tami Moser [00:04:17]:
The advanced implementation level 3 is where you go full integration with your e r EHR EHR systems. And what I've seen over time is that, you know, EHRs in general are built well, software in general is built to where you have different levels of that software. There can be add ons. Each add on cost you something extra per person using it, and your facility makes choices often based on financial concerns as to what level of package they purchase and how many add ons are available. So at this stage, what you might find when you get into discussions is that your EHR system has a tool, if you will, that could easily be adapted for what you need or specifically designed for what you're doing. But at this stage, perhaps your organization hasn't purchased that additional add on for the package. So this is something you need to consider. Right? But what it gives you is automated real time data collection, AI powered analytics.

Tami Moser [00:05:24]:
And for those of you who might be against AI, let me just step in and say that this is not going away and it's becoming more and more useful. You can really tap into the power of AI to help you in many areas. And so AI powered analytics actually can take this one step forward where it can make suggestions about changes that could be made. It can provide real time insights from the raw data that exists in there. So it is a tool that you could find very helpful. And many software programs now are building and trained AI to their software packages so that it works within the confines of not only a software, but the types of needs that people using the software would have. Just something to think about. You can also create custom dashboards, have mobile accessibility, so on your iPads and your iPhones, of course, security is an issue there too, so you need to make sure that it's secure.

Tami Moser [00:06:35]:
But it's best for large programs, or complex initiatives where you have maybe a lot of metrics that you need you absolutely do need to track and you need to monitor real time, and you need to be able to adapt based on what's found in those. So the cost can be significant here, again, depending on what you already have. And time to implement is months to a year, again, depending on what already is implemented and integrated and how far you're gonna go with this particular edition. Let's talk about HIPAA compliance and dashboard design for a moment. And here are some of the key considerations for this, data de identification. So you need to remove all 18 HIPAA identifiers, use aggregated data whenever possible, and implement minimum necessary access. In other words, you don't just allow everybody to look at everything. You actually define, which for those of you who are used to working with EHR systems, this is something you're already aware of or have had some exposure to.

Tami Moser [00:07:41]:
Security measures are the second here. So role based access control becomes crucial. So decide what roles people are playing in your program and then backing up and going what access to data do those people need. Some just need to be able to enter data per patient. Some need to have a full access to the dashboards and reporting system. You need to decide what that looks like. Encryption at rest and in transit becomes some this is where your tech people become incredibly helpful to this. Audit trails also.

Tami Moser [00:08:16]:
Who touches the data? What do they do with the data when they're there? And then secure login protocols. And I would also say there needs to be with the audit trails, it's kind of like the other types of metrics you're looking at. How often are you gonna look at an audit trail or even just do random checks of audit trails? And then 3 is physical safeguards. Right? So secure workstation, screen time out policies, printer facts considerations. I don't think we need to go to any more of this. This is also something that all of you that deal with health care information are very familiar with. But just as a reminder, for the development of your program. Now let's expand our discussion of feedback loops a little bit.

Tami Moser [00:08:58]:
Let's kind of talk about what a comprehensive feedback system would include, and there's 4 primary areas. The first are data collection points. Patient satisfaction surveys, clinical outcomes, staff feedback, resource utilization metrics, cost effectiveness measures. Now this gives you some very distinct areas in which you can collect data and have those points available. So look back at what you have already designed in outcome measures, balancing measures, structural measures. Look at your outcome measures, balancing measures, structural measures. Look at your process measures, then look at the measures you created from the last podcast to actually start getting a better feedback system for changing things. And now let's look at where are those data collection points? What what are they? And how many do you have in terms of patient satisfaction surveys, clinical outcomes, staff feedback, resource utilization metrics, cost effectiveness measures? So do all of those exist in what you have? And what you might have effectiveness measures.

Tami Moser [00:09:54]:
So do all of those exist in what you have? And what you might find missing more than anything is cost effectiveness measures. But having some data collection points in all five of those can be very helpful. The second is analysis methods. So we've already talked about this a little bit. Right? Statistical trend analysis and doing statistical analysis in general for correlation, pattern recognition, benchmark comparisons. So do we have a benchmark to compare to? If you are replicating a design that has been done elsewhere and their outcomes and different measures have been reported on, they may be the benchmark, right, to try to get to that level of, outcomes or success. And then if there are problems, root cause analysis becomes something that can be very useful that I don't believe has been mentioned in this at all. And so, well, maybe I've I've got an inkling that we have talked about it at least once.

Tami Moser [00:10:59]:
But root cause analysis is basically going, I don't want to just treat symptoms. I want the actual root cause to be cleared up. And so this can be related to issues associated with what you're doing. It can be problems that you've identified through looking at your data, and now you need to figure out what the root cause of that particular issue is. And so there are some methodologies associated with that, and you can really look into, okay, how can we get to the root? And I always explain it this way. Right? If I come in to you as a practitioner provider and I have a headache and you say take 2 aspirin every so many hours, and it'll be fine. You give me something for the headache, but you don't actually do anything beyond that. And the headache's just a symptom of a brain tumor.

Tami Moser [00:11:52]:
Well, I would much prefer that you actually do some analysis, do some evaluation, and figure out I've got the brain tumor so we can fix that versus just sending me home with something that will actually alleviate some of the pain of the headache, which is just a symptom. So it's the same thing in our organizational life and what is happening there. When something goes wrong, I don't want just a Patsy to blame or an easy out or a quick fix based on symptom. I wanna actually spend the time necessary to get to the root cause so I can fix that, and I don't have to deal with this again. So keep that in mind. I mean, that's the value of doing that level of analysis. The third is insight generation processes. So regular review meetings.

Tami Moser [00:12:41]:
How often are you gonna get together as a team and actually go through all of your data and really start looking at what is happening. What are we concerned about? What do we need to adapt to? You know, that data driven decision making for your team. And, really, you think about when you get together as a team and you look at reporting, what you're trying to do from that those reports is gain insights into what is happening. And then based on those insights, you adapt. Right? And so that's what the regular review meetings help you do is get to that point. Also, what data visualization tools do you have? Now I gave you you have links to 3 different books, and that is when we were talking about data analysis earlier in our program designing process. And going back to those can be very helpful now because dashboards, reading a rep reading data, even a reporting where it just gives me aggregate data, does not necessarily mean that has been turned into information. And if it's not information yet, in other words, I haven't taken it from data and put that data in the realm of context, then I'm not sure what I can do with that yet.

Tami Moser [00:14:03]:
Right? So once it's in context, I've got information to work with, and then that information can help us gain insights. And then those insights can go into decision making for us. And that I said to say, the data visualization tools are very helpful for people because often we need to see the visual to better understand what's happening with the data. And some people, as we all know, do better with just the data in front of them, and others do better by having it in a visual format. So don't overlook that, and you can go back to those other materials to help you get a better idea of how you could set up visualization. And that visualization, depending on how you want to create the visual, is a good thing to put on your dashboard usually. Get stakeholder input, and how are you gonna do that? So stakeholder input sessions. You go back to the stakeholder guide you put together in terms of how you communicate, what types of communication you do with them, when they're important to communicate with.

Tami Moser [00:15:10]:
And now you can take that over here into the insight generation process and go, we want insight from stakeholders on this data that we have. So how are you going to design those sessions? Who are you gonna pull in? How long are they gonna go, and what are gonna talk about? And then expert consultation. Do you need someone who's an expert in the area that you're working within to help you interpret and gain insight from that data? And so that that could be an expert that comes in because they're an expert in facilitating sessions for gaining insight. Right? Helping you move through the data in a systematic way and have conversations about the data moving toward insights and helping you capture that. The 4th is action implementation, having an action implementation framework. This is where you have your decision making protocols defined. So how do we decide what we're shifting, what we're not, what insights we're acting on? You need decision making protocols. Who's responsible for making decisions and what roles will they be in and what realms are they allowed to make decisions in? That kind of information needs to be defined.

Tami Moser [00:16:32]:
Then you need change management procedures. I mean, I got I teach classes in change management. It's a very broad area. People get specializations in their PhDs in change management procedures. But we have a lot of change management frames that can allow you or help guide you through changing. And so putting together a procedure on how you want to approach change in your program is an add on that is really important for your decision making protocols and an action implementation framework. Then your communication strategies. So we've talked about this in other areas of the program design, but this is where you kinda pull it back into the framework overall.

Tami Moser [00:17:23]:
You have communication strategies for your stakeholders or you should. You should have now defined communication strategies for your audience, your participants. Right? And now let's pull that all together in a more cohesive picture, action implementation framework for communication with timelines, guidelines, protocols, channels, all of that kind of needs to come together here. And then you're gonna monitor your progress, and that's a part of this framework too. So let's look at a real world example of this system in action with our diabetes management program as the example. So for data collection, here is how we're handling that, and these are our data collection points. So daily, there's blood glucose readings and medication adherence. So that's a daily data point.

Tami Moser [00:18:25]:
Weekly, we have an exercise log and dietary compliance. So we have those 2 data points that come in weekly. We have 3 data points that we gain monthly, weight, blood pressure and a one c levels. And then quarterly, we're gonna do quality of life assessments. And so we're gonna have all these data points coming in to our diabetes management program. For our HIPAA compliant dashboard elements in this program, we're gonna do aggregate 10 trends, so we'll have no individual data available in the dashboard. Population health metrics, community health metrics, program participation rates, and outcome measurements. So we're gonna go back to the measurements that we've defined as outcome measures, and we'll also have those on this HIPAA compliant dashboard.

Tami Moser [00:19:18]:
Then for our feedback loop pardon me. 1st, our data shows declining medication adherence. So we're seeing that on our dashboard. 2nd, our analysis reveals transportation is a barrier, and that's why we are having declining medication adherence. Our team generates solution options for this, and then we're gonna implement a medication delivery service. This is what we've decided is going to be the best shift or adaptation based on the insight that transportation is a very real barrier. And then we're gonna monitor impact on adherence rates. So that's gonna be the next part of what we're doing.

Tami Moser [00:20:05]:
Sorry. Take a quick drink, and maybe that will help. Let's discuss program adjustment strategies in more detail. There are several approaches, each each with its own advantage. So rapid cycle improvement. With this approach, we would have weekly data review, small targeted change. We would quickly evaluate those changes, and then this would be a fast iterative iteration. We would just go through these small targeted changes to improve on a regular basis, and it's a rapid cycle because it's weekly.

Tami Moser [00:20:47]:
Then we've got the a b testing methodology, which means we have a control group maintenance. We have clear success metrics. We have statistical validation and structured comparison here. So we've got our group a and group b that are going through the program. Then we have predictive modeling. Methods or strategies for a program adjustment. They're not the methods or strategies for a program adjustment. They're not the only ones available.

Tami Moser [00:21:26]:
You may have one in your particular organization that you use and really works well, and that's what you wanna plug in here, and that is fine. Or you may wanna use one of these. My my personal favorite, in other words, what I would default to if it works is the rapid cycle improvement. I would use a PDSA methodology, and we would do this on a regular basis through our weekly data review. So when we would sit down to have our weekly meeting and we would look at our data to gain insights, we would take an insight that we feel like is the most important for us to deal with right now, and then we would take a small targeted action related to that that changes what we're doing. And then we would evaluate that and decide where we're going from there. So that is my personal default, but that doesn't mean it's the only one available to you or that it's always the right one to choose. So it really is program dependent, and, also, it can be dependent on the skills that you have on your team.

Tami Moser [00:22:29]:
You wanna make it something that you can all engage in and feel comfortable with so that you you really do engage in it. Here's a suggested timeline for implementing your adaptive monitoring system. And again, this would really just depend on what you choose, but I felt like an example could be very useful here. So let's say month 1, we have stakeholder meetings. We gather system requirements, and then we do a HIPAA compliance review of our monitoring system. That's what we're doing. We're we're looking at the adaptive monitoring system here. Month 2, we would select our tools and have development of those tools.

Tami Moser [00:23:08]:
So if we're gonna use Excel, that's where we go. Okay. We're gonna use Excel, and this is how we're gonna develop that, or it could be a different tool. You'll do the initial setup and then staff training. And as we all know, staff training on technology can be a real cluster in many ways because people have very interesting relationships with technology, and comfort levels can vary greatly. And while we all tend to use it, I mean, it's integrated into our daily life, it doesn't mean everybody's comfortable in learning new technology. So staff training, can can be challenging, and that's why in month 2, you address it, and it may take the whole month to do that. In month 3, you're gonna pilot test your system.

Tami Moser [00:23:57]:
You're gonna do feedback collection, and then you're gonna refine the system. So remember, this isn't about your program itself. You're actually now testing your adaptive monitoring system to make sure it works and refining the system. And then in month 4, it's full implementation, ongoing monitoring, regular updates. Now, I put this in a 4 month format, if you will, because each of these areas may take a full month depending upon the level of complexity of what you're choosing for an adaptive monitoring system. If you're going with the beginner level easiest excel or Google spreadsheets, then it could take you 4 weeks instead of 4 months. It it really is dependent here, but going through all these steps is important regardless. So keep that in mind.

Tami Moser [00:24:54]:
Timeline, I mean, setting a timeline to it is also important. Right? Because you're planning out what you're gonna do and in what stages it will happen. So depending on the complexity of your adaptive monitoring system, the timeline for going through those steps will vary. So your work for this part has actually three parts. The first is to draft an adaptive monitoring plan. So define your metrics, define the collection methods, outline security measures, and describe feedback mechanisms. Now before we go to the next two components to this, what I want you to think about here is you've already done a lot of this work. Right? You already have metrics.

Tami Moser [00:25:40]:
You've already defined how you're gonna collect them or you should have by this stage. Security measures outlined to the best of your ability. You may want to go have a conversation with some of the people we defined earlier that can be important to you being able to set up a system and make sure that it is HIPAA compliant. So that would be part of what you're doing in this first component. And then your feedback mechanisms, how are you bringing back the information and then gaining insight and moving forward? That's kind of your feedback component. Next, we go to stakeholder analysis. You've already done this to a great degree. Now you're bringing it to this end stage and going, okay.

Tami Moser [00:26:23]:
Is this what it needs to be? Based on the information I've gained since we did that originally, what might change? So you should have identified key partners. You may add some partners now based on what we've talked about. Define roles and responsibilities within those stakeholders. So do they have any responsibilities? Are they engaging in the program or kind of an external stakeholder that more is more dependent upon gather gaining information from you, but not actually engaging directly with what's happening. Define roles and responsibilities. Validate that what you've already done is what still makes sense, or do you need to change anything? And then create the communication plan. So, again, review what you had already planned out and expand on that as needed. The third is implementation timelines.

Tami Moser [00:27:17]:
So, again, not completely new, but you need to think about it now in the broader context of everything we've talked about. Map out those key milestones where it's important for you to really take a stop to look, consider, evaluate and shift if needed. Okay? Allocate resources or discover where resources are and start the process for gaining access to those resources. So where are they gonna come from? And when you get them, where do they go? Who are you allocating them to? To what degree are they gaining access? And, again, as we all know, resources aren't just money. Money is a part of it, but people, materials, technology, how are we gaining access? Where are we allocating it? So maybe I'm running this program in multiple locations like our Healthy Kids Millbrook, and some are smaller locations, some are larger locations. So the resources allocated would obviously follow the needs of the size of those groups. And then set your review points. Absolutely where are you reviewing and to what level of review are you doing.

Tami Moser [00:28:40]:
So we may have our meek weekly meetings that feeds into our rapid cycle improvement. So that's how it would probably work for me. We have weekly meetings. We look at our data points. We look to gain insights from them. We define the most important insight that needs us to address it. Then we look at what that insight tells us. We do more research.

Tami Moser [00:29:04]:
We discover the underlying root causes of what we see, and then we do a rapid cycle small change to test it to see if that fixes it while not upsetting anything else that we have in place that's working well and those metrics. Right? So that would be probably that start of that review point. But as we know from earlier examples, let's say I've got some things that data points that come in weekly, monthly, and quarterly. So I may set up this program to go every quarter. We do a full review of everything, looking at all data and making decisions that could shift larger portions of the program. So while we're doing rapid cycle improvement at a weekly level, every quarter, we may have something larger that we focus on that needs to be changed. So that's an example of how you could set review points. Right? And and I could go at the quarterly.

Tami Moser [00:30:07]:
There's gonna be more people in attendance because it's gonna have more of the stakeholders involved at those quarterly meetings than in the weekly. Remember, the perfect system is the enemy of a good system. In other words, we often seek to gain or achieve a perfect system. And by seeking that, we give up the ability to have a good a fundamentally good system that gains us what we need. And instead, we focus on trying to get to the perfect system which almost never exists. So start small, ensure compliance, and build up gradually. Focus on collecting meaningful data that can actually be used to improve your program. And by meaningful data, I don't know about any of you here, but from my experience, I have sat in many assessment meetings where we look at data that is being collected, and some of it we're required to collect, and we can't do anything with it.

Tami Moser [00:31:10]:
It's not meaningful to us. It can't be used to to understand what's happening and what we and give us insight into what we could change to make it better. Therefore, it's not meaningful. If I can't use it to make decisions with, it doesn't provide us with information about success and overall outcomes. Why are you taking the time and wasting the energy on collecting it? And just because you can, doesn't mean it's useful. So, really think about each data point you defined. I would now go back and look at that and go, how is that meaningful to me? And it would even be good to set up a spreadsheet and have a column where you put meaning, which in fact, look in your Excel workbook. I'll put something there so that you can kind of outline this and put all these components in place.

Tami Moser [00:32:12]:
In the next podcast, we'll start looking at adaptive monitoring plan to stakeholders and integrating it into your overall program design. We'll also look at some success stories and think through implementation of these systems. Thank you for joining us for this extended session on implementing adaptive monitoring systems in health management program design. I'm doctor Tami Moser reminding you that good lit data leads to better decisions, and better decisions lead to healthier communities.