How to reduce workload in healthcare with smart algorithms

How to reduce workload in healthcare with smart algorithms

Consultation hours are overflowing and much of the time is spent ticking off mandatory lists. It’s driving caregivers crazy! Understandable, but not necessary. With a different focus and approach, there is a lot to gain in terms of time, money and satisfaction. The secret lies in looking at risks rather than diseases and using their own data intelligently with the help of algorithms. In this blog we describe how you can organize your care based on patient groups with a low, increased, and high risk profile. We show that this leads to: 1) better outcomes for patients, 2) lower work pressure and greater job satisfaction for caregivers and 3) more efficient operations and cost reduction for healthcare administrators. And the best part is: it’s within easy reach.

Focus on risk, not disease

The major pitfall of our current healthcare system is the focus on disease. These days, we are dealing with a growing group of people with two or more illnesses at the same time, who are at risk of falling between two stools because of this focus in terms of care. It is much smarter to coordinate and organise the provision of care on the basis of risks and the associated health needs of patients. By identifying unique subgroups of patients, healthcare organizations can provide the right care tailored to them rather than ‘one size fits all’. Thus, they have more influence on better outcomes for these patient groups. In addition, they can organize care more efficiently and cost-effectively. This means that the traditional classification (diabetes, COPD, heart failure, oncology, frail elderly) of groups will become obsolete. This is a new way of thinking, at least for the Netherlands. In other countries, such as America, Australia, New Zealand and Colombia it has already proven its effect. Best practices there show that realizing the right care in the right place starts with segmenting patients based on risk stratification models.

How does risk stratification work?

All healthcare providers must record data. GPs do this in the GP information system (HIS) and hospitals use a hospital information system (HIS). By applying algorithms to the data in such a system, patients can be categorised fairly easily on the basis of risk profiles (see Figure 1). Some people have trouble using algorithms because of negative associations. This is unfortunate and unjustified, because algorithms can detect patterns much better than an average healthcare provider, andthis will benefit your organisation enormously. What exactly, we will try to explain by means of three risk profiles that can be detected by means of risk stratification. We will briefly outline the profiles and then indicate step-by-step how you, as a care provider/institution, can go about tailoring your care provision to these profiles.

Figure 1: How risk stratification works

High risk patients

This group, which represents approximately 5% of the population, includes people who, for various reasons, visit the family doctor and/or another healthcare provider more than twenty times a year. The superusers , so to speak. As expected, the frail elderly are well represented in this group. We also find the Wajongers with mental health problems there. But the group is more heterogeneous than we think. Risk analyses show that middle-aged women with families are also superusers. A group that does not easily attract attention, but – just like the other high-risk patients – can cause clogging up of the consultation hour in daily practice.

• Step 1:

  Pair each high-risk patient with a case manager. This may be a nurse practitioner, a practice support worker, a district nurse, a GP or a geriatric specialist. The Case Manager coordinates a patient’s care across the continuum of prevention, care and well-being.

• Step 2:
  

  It is important that the case manager gets off to a good start with the patient. For example by making an appointment to get acquainted and to tell about his/her role.

• Step 3:
  

  The case manager then draws up an individual care plan together with the patient. For this purpose, a standardized approach is available with fixed templates and documentation possibilities within the electronic health record (EPD). Such a care plan addresses clinical and non-clinical problems. Self-care and patient involvement are always part of a care plan.

• Step 4:
  

  Since a high-risk patient almost always has to deal with both care-related and non-care-related problems, it is important that the Case Manager has a broad social map at his/her disposal and is able to cooperate well.

Increased risk patients

This group includes people with one or more chronic diseases that alternate between stable and unstable periods. It is a good idea to include in the risk stratification factors such as obesity, depression, high blood pressure, high cholesterol, elevated blood sugar levels (pre-diabetic) and tobacco use.

The aim of care provision must be to ensure that these people do not move on to the group of high-risk patients.

• Step 1:
  

  Coordinate care from a care group or health center in which all relevant disciplines are represented. Foreign models such as Accountable Care Organization, and Patient Centered Medical Home (PCMH) are examples that use this methodology.

• Step 2:

  Try to involve patients in the care as much as possible. This can be done through self-help support, shared decision-making, written care plans and patient surveys. Family and friends can also be involved asinfluencers. They can stimulate the patient in self-care and also act as an extension of the care team, for example by informing the care team when important changes occur.

• Step 3:

  Schedule annual preventive screening moments for chronic conditions. This allows gaps in care to be quickly identified. A health coach can also be used for this.

• Step 4:

  Invite high-risk patients to try new forms of care. Examples include eHealth applications, online forums, group meetings and health coaches. Research shows that it is also good to involve friends and family in this.

Low risk patients

This is the group that remains: the people who are healthy or stable in health. The goal of care is to keep these people healthy. This can be done in a cheaper, more efficient and more effective way than is often the case nowadays.

• Step 1:

  Develop an online patient portal for initial communication and encourage low-risk patients to use it (Figure 2). On such a portal, information can be put about the most common disorders and what to do in that case. This stimulates self-care. Research shows that this way of working increases patient involvement with the organisation. This is interesting because it can stimulate efficiency in the use of care. For example, the number of consultation visits can be reduced.

• Step 2:

  Focus on low-risk interventions. As described above, it is important to efficiently manage the population of low-risk patients. Look for other strategies to better reach this target group and prevent them from having an increased health risk.

• Step 3:

  Reserve time and capacity for (online) prevention and avoidance of unexpected care needs. This does not always have to be done by a doctor; often care provided by a practice support worker or doctor’s assistants is also sufficient.

Figure 2: Online | eHealth Interventions

So what does that get you?

Research shows that risk stratification is the way to put value-based healthcare (VBHC and/or Triple Aim) into practice and actually realize better outcomes. It forms the basis for proactively providing tailored care. Coordinated, person-centred care saves the patient unnecessary treatments and therefore time. In addition, with a focus on prevention, a lot of misery can be prevented. All this leads to higher patient satisfaction and better health outcomes. Care providers will also reap the benefits. By organising care around high-risk groups, valuable time can be used much more efficiently. Instead of scheduling ten minutes for each consultation, make a distinction between the risk groups. You give low-risk patients less time and high-risk patients more, which means you can help the latter group better and will probably see them less often at your surgery. In addition, tasks can be better distributed. Not every patient needs to see a doctor for every problem. In fact, some matters can also be handled perfectly well online. All this results in a lower workload, lower costs and more job satisfaction. Finally, risk stratification enables managers to make more targeted contractual agreements with health insurers. These agreements are currently based on volume and price, with national figures serving as the starting point. But there is no national average. The need for care in Amsterdam is completely different from that in East Groningen or South Limburg. If you map this out, you can conclude contracts that are tailored to the regional situation. This improves the quality of care and saves costs.

But what’s stopping us?

That is indeed the big question, because it does not really want to get off the ground. There is still too much talk and too little action. Risk stratification is only possible if healthcare providers have their data in order and report the correct data. That is, not process indicators, but outcome indicators, such as quality, health and cost. It is also important that data systems are linked. And don’t say you can’t because of privacy reasons. That’s a fallacy. It is perfectly possible to link datasets and comply with privacy legislation. Technology and legislation are therefore not the stumbling block, it is the prevailing culture. It seems like we have some sort of fear of the unknown. We should really dare to step over the line. As healthcare providers, we do have some influence on the overcrowded waiting room. We can negotiate better contracts. Let’s try and be proactive about that. Instead of waiting for the government or health insurers to take action when the situation becomes untenable, take matters into your own hands and start working with risk stratification. In doing so, we stand up for ourselves and for the patient.

Want to know more?

Want to know more about how to use smart algorithms in practice? And what else is needed to make data management & risk stratification a success? Download our e-Book Care networks that work.