06-13-12 | Blog Post
I’m liveblogging from the iHT2 Health IT Summit conference in Fort Lauderdale from June 12-13, and Online Tech will be sponsoring and exhibiting our HIPAA hosting solutions as well.
The conference focuses on healthcare IT as the industry evolves, including the latest implications of ARRA (American Recovery and Reinvestment Act of 2009) and meaningful use changes and how it affects practices and hospitals. Other presentations will focus on electronic protected health information (ePHI) security, using EHR systems for health information exchange (HIE), healthcare delivery reform and other topics on improving patient care with data technology.
Panel “Connecting Points of Care: Leveraging EHRs to Make HIE Work”
While the debate economic feasibility is still underway, numerous case-studies and pilot projects have demonstrated the successful implementation of health IT programs and the significant benefits to outcomes, safety, and healthcare costs. In order to maximize the benefits on a national level, healthcare stakeholders must begin sharing this information – in real time. While there is significant work to be done before all providers are able to share information in this manner and on a national level, there are many steps that can be taken now to improve communication and coordination of care on a local and regional level.
This session will discuss:
Getting started with HIE
Moderator: Jean DerGurahian, Executive Editor, SearchHealthIT.com, SearchHealthIT.com
Leslie Durham, Chief Information Officer, Bethesda Memorial Hospital
Richard Biehl, Ph.D., Adjunct Professor, Department of Health Management & Informatics,UCF College of Health & Public Affairs
Shadaab Kanwal, MBA, MISM, Executive Director Research & Quality,
Kaiser Permanente – iHT² Advisory Board Member
Jayne Bassler, VP & CIO, Florida Hospital – iHT² Advisory Board Member
Leslie Durham: Of the 17 hospitals in Palm Beach County, only 3 are non-profit, so it’s very difficult w/ the corporate facilities to figure out how to share patient information. I need to share information with our competitors down the street to know if there are patients going from ER to ER shopping for meds or are coming with a pre-existing condition.
Jayne Bassler: Our 2 CEOs came together and agreed that for the benefit of the community, we had to share patient information. While we hold true to our commitment to community benefit, the model changed from a technical perspective. Originally we were going with a federated model, but now we have a repository model. As a large systems with data available for research, we need to figure out usage of the data. It’s primarily in the ER departments with a 75-85% hit rate from physicians adopting the system. But we’ve made it very easy for them. There’s no sustainability model in place to keep this thing alive after the initial investments. We’re starting to let other providers into the system, but they can’t contribute yet because we don’t want to fund the entire project.
It’s going to be an interesting time as we make this transition to get providers the information they need and still provide the community benefit.
DerGurahian: Comment on trust and data integrity?
Biehl: Start with data governance. The difficulty of building HIEs should humble us. Try connecting 75 systems within an institution and if it’s easy, I hope you’ll let meknow. It’s daunting. A medication allergy to apples and oranges instead of a normal medication is a challenge because it’s not a medicine allergy. It is an allergy to both apples and oranges separately, or only when apples and oranges are eaten together. I don’t know how to interpret “severe rash”. Putting governance in place and getting true definitions is key. Having good control over the data is a prerequisite. Do people receiving the data trust the data enough to really use it effectively at the other end?
Kanwal: We’re talking about patients and their families, so we need to extend the trust ot the public environment. KP exchanges data between multiple organizations and these exercises happen as soon as a patient shows up i one of these organizations. Data quality issues are a continuous effort that starts at the point of entry and continues with clean up along the whole path.
DerGurahian: Is there a way to bring the systems together? Where do we go with this?
Kanwal: It’s a function of knowing what we will allow to share. If I’m a terminal patient, I may not care as much about who sees my data because I really need the care. Situationally, it might depend what part of the system I’m involved in.
Bassler: Speaking from my background as a nurse, we’ve been exchanging information the whole time, and sometimes we fall victim to freezing in the conversation of ideal information exchange instead of moving forward even though it’s not perfect.
Durham: We still have to look at the data as we would in a paper chart and use our background and skills to interpret that to patient care.
Biehl: We’re going to run out of eyes to look at paper charts. People can recognize ambiguous data, for example, seeing a temperature of 986 instead of 98.6 is going to quickly tell a provider that there’s a decimal point error, not imminent patient incineration. But this depends on human eyes and brain to make this interpretation. In the future, when there aren’t enough eyes, how am I going to address this issue. RIght now we’re saved because we have this knowledgeable clinician base. When computers start looking at this data, we’re going to see some real challenges.
Kanwal: These are issues where we look to technology to solve the issue. But the problem is a systematic integration. In comparison with a systematic application of standards, an integration without applied standards is not nearly as successful.
Bassler: Our main objective was patient benefit so physician perspective was key. I want to see us progress and not let each individual patient be sacrificed in the interest of perfection.
Biehl: In a federated model, I tend to hear about it from the users instead of the IT support team.
Repositories keeps the burden on the IT department instead of the users.
Audience: I was asking for a patient what type of meds they are getting, I had to translate the terminology, so it goes both ways. Any insights on sustainability models for HIEs would be appreciated.
Bassler: We haven’t figured it out yet. We are looking to the payers to help. As we’re bringing more people to the network, we’re starting to bring on more of the underserved population, and we’re trying to figure out a model that doesn’t put a burden on them, but we’re trying to figure out a balance.
Audience: If you can come up with an answer, please publish it.
Biehl: I don’t know how to implement the idea, but an iTunes model with a fee for service based on data usage where institutions, and payees, and patients could realize financial benefits at some point down the road, it would be sustainable.
Audience: What are the thought processes to flow information to other entities.
Bassler: Since we only share information with 2 entities, there’s a coordinate communication with the vendor., The problem is ensuring the timeliness of that happening. THe hospitals are 24×7 and some of our vendors are not. It has to be real-time changes to make sure that patient care is appropriate.
Audience: Is there a feedback loop to fix errors in the patient record?
Bassler: We have not figured that out yet. While we know that 75-85% of the physicians are looking at the data, we don’t have any statistics about how they are using the data. Early on, we were trying to understand how to capture errors, remedy them, and be able to prove benefit. There was a risk-adverse reaction for physicians to document what they did or did not do with the information for fear of unknown negative consequences down the road.
Biehl: We’re working to assign logistics models to healthcare models. We’re using lot quality control which is a mature field in manufacturing logistics to see how this might help in healthcare. The HIEs being built today don’t have that functionality today at all – it’s an added complexity. It will be hard to build if we don’t document the need.
Bassler: As new contributors come into the HIE, how does the change control work and whose responsibility is it to test and fix the data. We don’t know how to address and fund this.
Audience: A lot of HIEs cropped up because “we had to have an HIE” but there was no value proposition in the first place to be able to measure and prove value for the investment. You have to figure out why are you doing this before you start.
Durham: That’s been one of the big problems starting this in Palm Beach County. We understand the value, and that’s not as hard to prove as opposed to sustainability.
Audience: Eye images: currently there are no standards to transfer them around. In an HIE where you can attach images and pass them around, that got rid of courier services and the participants were willing to pay for this service.
Audience: Would it make more sense to use an interactive information network instead of a repository?
Bassler: That was the original intent, but there were complications in the vendor delivery. The executive team needs to stay close to the process and once we got into a technical discussion, decisions were made that the sponsors were unaware of and we ended up in a different place than we originally headed. If we have to move to a system where providers have to check for new information without knowing if the new information is there yet or not, we will expect to see our usage drop.
Biehl: This is where we expect federated information systems to mature. We all have to agree to be much more proactive about releasing data. We’re not there yet, but eventually “the cloud” should get us there. I don’t see our federated tools being there yet. [Reaction to audience question] The way we spend money in-house instead of out-the-door is very different. We’re happy to spend $5 in-house over $.02 out-the-door, but if we bring these costs closer to the physician, we might see a different reaction and decision making process at the point of care.
Audience: We have incentives for providers to keep the system of record in place. If they don’t maintain the information in the system of record, we don’t want to work with those providers.
Audience: What’s the appropriate role of the government in the development of information exchange?
Biehl: We’re a very entrepreneurial culture, and I don’t see a problem with funding. If they want to get involved in how we manage our patient data, that would be a concern. We have to realize that this is a global problem, and not going to be solved by government alone.
DerGurahian: Does it help or hinder to have governance in the structure of exchanges?
Bassler: The government is the biggest payer, so it makes sense that they are involved in the conversation, but if they created and enforced the structure of information exchange, that might stifle innovation and progress.
Biehl: We can push HIE out faster than we know how to deal with it like genomic or proteomic data that we don’t have any idea what to do with it, but we need to focus on the usage and interpretation of the data at the other end.
Biehl: We have to want the movement of data down the pipe and the impact we have on the clinician by sending it. I’d like to see more research into how the information entered by the physician changes when they know it’s going to be shared. Will there be less candor? We’re demotivated the physicians to share information that they wouldn’t want the patient to see and we may be losing a nuance of care.
Durham: We’re also need to consider what information the patients want to share. They may not want mental health information to be shared with all members of their care team.
Town Hall “Health Information Exchange Strategies – South Florida”
 |
 |
 |
|
Executive Director South Florida REC |
SVP & Chief Information Officer Memorial Healthcare System |
Chief of Informatics Harris |
Forest Blanton:
Memorial is funding the HIE project. We have a very low cost of entry ($120 yr) and the cost is lower for larger group practices per doctor.
If other health systems join, they need to be responsible for supporting their doctor’s IT needs.
Our governance model is being revised to make it more participative and open.
Lessons learned:
Be as transparent as possible
Be patient – it take a lot of perseverance to work through all of the contracts. Most of the issues have not been technical, but more contractual.
Promote the value
Dave Riley:
To Dr. Biehls point about self-censoring in the record, medical records are undergoing a transformation from a personal record and place to jog the memory to an act of interpersonal communication. This act of censorship is the realization that it’s a more public record. I don’t think this is a negative, even though the linguistics around the communication will change.
With respect to the value of the HIE, and when you form networks (phones, faxes, HIEs) as they go through their stages of gaining membership, the fundamental laws of the economics change. As soon as you get more member, Sarnoff’s (sp?) law kicks in – each new person adds value to the network. It makes a transition from additive to becoming multiplicative (Metcoff’s law – sp?) and then it becomes exponential (Reef’s law – sp?). There are a few research papers published about this and we’re beginning to apply these concepts to calculating the value of the network. There has to be a value to the people you’re asking to pay for it, and you have to be able to articulate that value. As I’ve travelled across the country, I’ve read almost all of the propositions in the grant applications. As I’m seeing boots on the ground approaches, it doesn’t match the reality of what we see at the national level in the rarified air of DC, so this reality testing phase is ok, even though it’s painful. We need the benefit of the lessons learned.
We tried to be as agnostic as possible with respect tot he model. We wanted to have APIs that didn’t care about the business of what was going on inside. The details of the structure didn’t make much difference to us at the national level. We needed to make sure people had local autonomy and local accountability as we were thinking about the structure.
We also looked at it as a network of networks. In my view, HIE is all about the transactions between organizations – referral of patients, transition of care, etc.
Geographic exchanges vs non-graphic exchanges.
Rawlins: Can you tell us about direct secure messaging?
Dave Riley (Harris): We use SMTP messaging for the secure exchange of email, What ONC was trying to do was create a simplified method of exchange using some standard technologies THe analogy was faxing from point A to point B so the need for really structured documents wasn’t as great. You don’t have to have an EHR to participate. EHRs are one user system, but if you use Word and can attach an email, you can engage. Just think of secure email and those use cases.
Audience: What is the value statement for Memorial for creating the exchange for the leadership?
Blanton: We didn’t define the value before we started. We looked at it like a pharmacy system or an elevator. Our physicians like it for increase productivity and we understand the values of reducing duplicate tests and having the data at hand when it’s needed. As all of us move to more integrated care networks and look at the the aspects of ACOs, HIEs take on another element to look at gaps in care, population that needs better health management. We’re not doing this at this time, but clearly the HIE sets the foundation to let this happen.
Audience: Where is the oppty for the patient to engage?
Blanton: Epic has a PHR (MyChart) that we use for patient engagement and we’re rolling it out to all hospital patients. There are a couple of ways patients can interact. THey can request a change in their consent. There is an availability for them to contribute information as well, but we’re not using this yet.
Riley: One of the built in functions of the HIE is auditing so a patient can see who access their record. We’re starting to look at what standards need to be in place to exchange data across the HIE for non-tethered EHRs. We’ll start to see much higher demand for patient engagement that will drive more of these standards.
Keynote Presentation “Leading Clinical Transformation…Think Different!”
Dr. Tayrien leads an expanding team of clinicians and informaticists responsible for the design, configuration, implementation and support of advanced health information technologies at HCA. The EHR team drives clinical transformation of HCA’s acute and ambulatory care facilities enabling improvement in efficiency, efficacy and quality in patient care by engaging physicians in transactional use of the technology. Parallel, Dr. Tayrien provides leadership for HITECH compliance assuring that HCA is on track to satisfy all Stage I “Meaningful Use” requirements across 162 hospitals.
Dr. Tayrien is also leading migration to the next generation electronic health record at HCA. Aligned with the organizations movement toward an integrated delivery model, the EHR, clinical data warehouse, and HIE will provide the foundation for patient-centric care coordination, evidence-based clinical-decision support, and innovation spawned through advanced clinical analytics.
Dr. Tayrien graduated with honors from Oklahoma State University, College of Osteopathic Medicine and completed his residency in Internal Medicine at OSU Medical Center in Tulsa. He is board-certified and a Fellow in the American College of Osteopathic Internists.
Chief Health Information Officer, VP, Clinical & Physician Services Group
Hospital Corporation of America
It’s the crazy ones who make the difference. Our goals are to make patient care:
Jen McDonough’s Story – a mom who trained for an completed an IronMan, but shortly thereafter her son was diagnosed w/ Type I Diabetes. She was passionate about finding the best care for her son, but struggled to figure out how to afford it. She said the process was SO draining, I had to go to the ER to get him in, so they had to ask me all the same questions again. In the pharmacy, they couldn’t read the Rx and wanted $500, and since then, I’ve been on the relentless bent to get all of the records from every lab and every doctor.
She and her husband had 10 W2s between the 2 of them trying to get out of debt and finance her son’s care. She didn’t have time to repeat her story over and over and over again and call to follow up on test results. But she didn’t’ want her son’s care compromised or to anger her providers who she genuinely appreciated.
How do we put patients at the center? How to we move to longitudinal care from episodic care, to coordinated care from fragmented care. When it comes to physican adoption, we’ve taken a volunteer approach and allowed docs to opt out and dragged it out way too long. It’s proven that it’s most efficient to get everyone on board as quickly as possible. We’ve invested about $1.9M in each hospital. To not take the 2 week go-live and get as many docs online as possible is a waste of resources. Wherever you are 30 days after go-live is about where you’ll be going forward unless there is a significant and costly 2nd campaign of adoption.
There’s a vast difference between online training and face-to-face, peer-to-peer communication.
When they studied the adoption of tetracycline in the hospital, they found exponential adoption in highly networked environments.
Attributes of adoption. We’re getting 70-80% adoption and we attribute it to a focus on investing in leadership training. At the end of the day, you need to be engaged in getting their input, looking at the communication channels in the social network? It’s related tot he influence of the individual docs to see who’s driving care in the hospital. A good surrogate for who’s driving care in the hospital is looking at the Rx order volume. Across 56 hospitals we’ve studied across HCA, 14% physicians account for 65% of the Rx orders.
We scale docs based on their Influence and Risk at the hospital. We choose influential docs who are supportive.
If you ask my team what I’m most passionate about, it’s about training at the front-end. If you’re not trained, you have no chance at adoption. 90% of the leadership docs driving order must be trained before go-live. We use a dashboard and track it daily to show who’s been trained, who showed up and didn’t, who needs to be trained. We want to make sure we’re on track to hit 90% before go-live because it’s too risky to scramble to train at go-live.
Our goal now is scale. We have a week-long boot camp to immerse the CEOs, leaders of quality management into the training process, and use a credible external source to help communicate the message.
These things don’t happen if the content isn’t correct, Referring to the Tipping Point, is the accumulation of hte little things that make a difference. You have to make sure your processes make sense to the docs and nurses, and if it doesn’t you have to match it up.
Example: 30 Steps from a written order to administration of a medication and best case takes about 2.1 hours with no time sitting on the unit coordinators desk.
Using an electronic CPOE eliminates 12 steps and reduces the time to 39 minutes or 300%. It also adds several checkpoints to check for errors.
Decoupling of Technology and Workflow
We created a virtual hospital for testing. We’ve fine-tuned it so that doctors don’t recognize it from the default implementation.
We’ve compressed the time from first engagement with a hospital to go-live to 7 months. Day 1 we have to present a model
“Training takes too long. We did not get enough training” Two direct quotes from the same physician pre and post go-live.
This is an opportunity to sit down one-on-one with physicians to understand their key clinical processes. This is a rare opportunity for hospitals to engage in this conversation and it can revolutionize a hospital.
Clinical Decision Support (CDS) is the holy-grail we’re all working to get to. There is a lot of opportunity, but many disappointments. Recent studies suggest that CDS is less than effective. John Hopkins found only 4% of alerts were accepted of over 40,000 alerts. VA Medical Center sat down w/ physicians to see what was going on. Examples are alerts for dual inhalers, dual narcotics, acetaminophen and aspirin which are all common place and do not present risk to the patients.
Theorem of Medical Informatics
(Clinician + Technology) > Clinician
The combination of Clinician + Technology should be more effective than Clinician alone – otherwise, why bother?!
The technology needs to increment knowledge and present new information or surprises. We don’t hit that bar very often.
Thinking differently about CDS. AlertSpace – web based tool that allows edits to pharmacy data that will persist. You can take some of those duplicate alerts and change their severity to reduce them to a more rational level.
Predictive analytics has some very exciting potential. We can take the current knowledge base and look at our own clinical data to create some predictive models and break out different cohorts to evaluate different treatment results. With visualization of pathways that lead to better or worse outcomes, you can eventually warn physicians if a patient is heading down a pathway that leads to a poor outcome.
Convergence of Technology and Medicine.
Book Recommendation: The Creative Destruction of Medicine, Eric Topol, MD
WorkForce issues. The biggest challenge we’ll have is the lack of healthcare IT individuals. We’ll need about 250,000. We’ve created our own HCA career page that are dedicated. We’ve created hCare University to hire college students directly out of college to put them into a boot camp for 8 weeks to train them in healthcare. It’s worked out great but now I’m on more social networks that I had ever imagined or hoped to be.
hcadoc.com is becoming a very important part of our implementation strategy.
We’re trying to implement all channels to increase our Health IT workforce instead of just stealing from across town.
1:15pm Panel “Accountable Care & Analytics: Driving Accountability through Information”
Regardless of the model for healthcare delivery reform, the end goal is for providers to efficiently and effectively manage care delivery. This includes providing high quality, safe care with few medical errors and strong clinical and financial outcomes. Although similar risk sharing schemes were tried in the past, none proved successful due to insufficient cost savings or poor clinical outcomes. For ACOs to survive this initial phase and become a legitimate model for the delivery of care, providers must become data driven organizations.
In this session participants will:
Explore different models for healthcare delivery reform
Review strategies to make providers data driven organizations
Learn how to apply ACO business intelligence techniques across the healthcare continuum
Moderator: Judy Hanover, Research Director, IDC Health Insights
Michael Brown, MD,MS, Chief Information Officer, Harvard University Health Services
John Santangelo, Director of Information Technology, Cleveland Clinic Florida – iHT² Advisory Board Member
We’re looking at our existing platforms and replacing or improving our front end tools. There are vendors with excellent technologies and others with a good content approach that know what structures relate the information, but eh connection between the two types of vendors has been hard to come by. We continue to use our own tools to analyze our data.
We’re looking aggressively at applying tools to improve our employee health and a competitive analysis of these platforms. I’ve been pretty impressed. The correlations and inferences these tools can make are very strong.
Hanover: How do you go about prioritizing analytics goals?
Brown: FIrst you have to decide as a culture what you want to accomplish, determine the workflows that will help meet those goals, and then the data required for those workflows. The first step is figuring out the priorities from the organizational perspective. A lot of our goals are operational, financial, and clinicial, with IT coming last. The insurance industry is doing some very impressive analysis and we’re also looking at some of those tools.
Hanover: The trend to self-service and being able to come to the data themselves is becoming important. Forest, you mentioned looking at many front-end tools – is self-service important part of the evaluation?
Blanton: We’ve done lot of pilots, primarily in the financial departments, and had a lot of trouble weaning the users off of tools that need to go away. We don’t want to risk throwing off the entire infrastructure to bring in new tools, so we use a lot of pilots.
Hanover: John, can you speak to the issues of unstructured data you are addressing?
Santangelo: This revolves around the ICD-10 issue to understand the clinical language to read the free text from the clinician and understanding the context where the message is coming from to understand intent and options that match the codes behind the scenes and presenting them to the clinician to make the selection. It’s a way to take the complexity of ICD-10 with analytics behind the scenes.
Hanover:
Blanton: We are not looking at an ACO pilot, but we are looking at best practices and techniques. We contract with actuaries to predict our future costs and we use this information as a starting point to inform our investment decisions. Some things we’re looking at is the ability to stratify and identify risks in the population. The projected health consumption patterns, predictive modeling that can take claims data and predict which members are more likely to consume additional resources in the future so we can look at ways to reduce those costs.
Looking at the diagnosis and claims data we can use evidence-based data to measure gaps in the care on a population and individual basis. On the financial side, people need tp know the cost of care. We need to know the cost of care outside of our systems in addition to just inside our system.
We need performance reporting to benchmark our reporting.
Hanover: John, can you tell us about the quality initiatives of your program?
We look at inefficiencies and qualities to take some of the cumbersome pieces out – this normally improves quality. You can lose quality if not done in an effective way. We need to optimize both the functionality and usability if we’re really going to improve outcomes.
Brown: For high-risk patients, you can do something simple like determining who hasn’t been seen in a while to touch base with them or on the other side, who has been in so often that they should have a case manager to work with them.
Some of the biggest opportunities and challenges are putting information in front of end-users. 🙂 It’s been a constant headache, but it allows people to think about what they are doing in a different way. The nature of the job has changed to answering the question “why can’t you get me this information?” It’s a very busy and frustrating ongoing process, but well worth it for the benefits.
Audience: You spoke to the benefits of natural language processing to improve unstructured data. Has this changed your approach to capturing unstructured data? It is more structured or does it allow more latitude.
Santangelo: IT’s definitely moved to the latter to allow more latitude to allow free-text.
Audience: Stepping beyond ICD-10, other areas of interest such as “Why can’t I get the information?” may come down to information being entered in the wrong field.
Santangelo: For us it’s a big challenge to get all the discrete fields possible while still allowing free text. We uncover many processes and workflows along the way. We don’t want to take away functionality they had before, so you have to try to get to data that isn’t in the EMR yet. You need to understand what’s being said in the EMR record and putting it in a reportable format. Physicians don’t care that their information wasn’t entered in a discrete field – they just want to access their data.
Brown: It’s amazing how complicated reports can be, and the more complicated they are, the less change people will actually use them. One the best justifications is not #MU, but the fact that we use that data. It’s a very clear message if data is not coded correctly. Sometime the reporting ability needs to be replaced at the source.
Audience: As new information came into the feeds, they became more discrete, so the problem list evolved to a more discrete list, so at some point the reports have to recognize a shift.
Brown: The reports can only be as accurate as the data.
Blanton: There’s an interesting tension between structured data and free text. On the one hand, some groups want complete codification and structuring of data, but this is a huge shock to clinical workflows.
Brown: I would not want to get in the want of good quality care by being too obsessed with codification. I like to read in the notes that include personalizations about the patient that can’t be coded.
Hanover: The data quality findings are interesting. In one piece, they looked at the smoking status in a discrete field vs free text and in that case, the data in the free text note was more accurate then in the discrete field.
Brown: Often physicians can describe a condition much better w/ free text than with codes.
Blanton: We have a group that manages both employee health, a portion of the Medicaid population, and an uninsured population. One of the things we always come around to is “where did the data come from?”. Sometimes data was collected from each physician’s office, but that’s difficult. Others use retrospective claims data and this is the direction we’re taking.
Audience: #MU2 produces some interesting results relating to certified reports.
Blanton: In our high-throughput reporting through Epic, but we’re not using the standard Epic workflow in our ED so we are using our own, non-certified reports. Now we’re at a point where we need to certify our own reports or go to the standard Epic workflow, and I’m not sure how to solve this right now.
Brown: I’m a little nervous about Stage 2 for automatic notifications because we know that our vendor product comes up with a result that we know is inaccurate. I submitted that question for #MU2 feedback to ask what we’re supposed to do in this situation.
Hanover: Analytics for operational and financial performance. What are your initiatives here?
Brown: They do more without my involvement now to choosing the plan rate, choosing preferred providers and deciding on our referral network. We’re being asked to answer questions like “What is our productivity in different areas?” or “What are our underutilized areas?”
Santangelo: Now we’re trying to take advantage of the front-end to optimize things like patient scheduling. If you can increase efficiencies and improve patient access to your services, to me this has a direct impact on the bottom line. On the front-end, it’s about optimizing the physician’s time and making it easier for patients to access our services.
Utilizing online scheduling to get around things that delay access to services. For example, in the past a surgical candidate would have to come in to fill out pre-op forms, but now they can do it all online from home. We’re trying to do a lot of things ahead of time that don’t have to happen during the visit.
Blanton: Our finance folks have been using analytics for years, they look at variances, trends, etc. Most of our operational questions float around labor. We’ve had a daily labor utilization report for years. Most recently, we’ve been looking at our staffing ratios. In the past, nurse managers were measured on level of compliance, but not penalized for being over-staffed. Now we’ve done a great amount of work with our staffing algorithms to determine when to round up or round down to improve our staffing ratios. We don’t use payroll data in our analytics, but we do put hours and over-time in.
Brown: As patients begin to engage in self-serve models, there’s a great way to be more efficient by measuring the opposite things. For example, don’t measure productivity to measure # of patient visits/day if you are reducing patient visits with online interaction. You have to be careful to make sure the measurements are a true reflection of your goals.
