Two years ago, Kingdom of Saudi Arabia’s Ministry of National Guard Health Affairs' Riyadh-based hospital, King Abdulaziz Medical City, became the first in the world to reach Stage 7 in four different HIMSS models. (It's recently become a pioneer with some impressive work to reach Stage 6 on another model.) Its advanced use of health information and technology has been a boon for the health system's 1.3 million patients.

Since then, the 3,720-bed MNGHA has continued its digital health transformation efforts across a variety of specific use cases, including an ostensible simple one that has long vexed provider organizations the world over: no-shows in outpatient settings.

They're disruptive, they add unnecessary cost to the care delivery process and they can have real effects on care management and patient outcomes.

But the Ministry of National Guard Health Affairs has been able to achieve some notable gains in reducing no-shows by applying artificial intelligence to its analytics, says Huda Al Ghamdi, director of data and business intelligence management at MNGHA, using AI to proactively predict which patients might be most likely to miss their appointments in ambulatory settings.

The health system is using machine learning to take data from its electronic health record – patient summaries, clinical information, appointment history – and process and train it for AI models that can alert physicians within the EHR – helping them send needed reminders to their patient and even booking appointments within their own workflows.

MNGHA comprises more than 30 hospitals, specialty hospitals and primary care centers across Saudi Arabia, with all facilities linked to a unified EHR system called BESTCare.

That gives the "advantage of having a huge amount of data," Al Ghamdi explained. "Advanced analytics, prediction and machine learning."

Innovative approaches to analytics have helped the health system in many areas, she said, but no-shows were a particular area of concern.

"The reason for tackling this problem in particular is because the outpatient setting is considered the biggest channel where MNGHA is providing the medical services to the patients," she said. "Unlike the inpatient or ER, outpatient is considered the biggest because we are talking about something like 20,000 visits per day [on] average."

That adds up to five to six million visits per year.

"So having a problem like a no show, it's definitely affecting the care providers, affecting the resources, affecting the patient themselves," said Al Ghamdi.

The fact that MNGHA is a governmental hospital means that sometimes it's difficult to measure the cost when patients don't show up for their appointments, she notes, but there is a cost, "and we should be aware of it and start thinking about saving."

Luckily, MNGHA has a "huge amount of data that we can start analyzing and studying and trying to figure out the factors affecting this," Al Ghamdi said. "We have a unified electronic medical record system that has different modules for registration, admission and outpatient.

"When it comes to the datasets we're utilizing in this project, it's mainly the demographic information, very simple information, mainly gender, age, in addition to the information related to the clinic itself, because there is a variation of no show from one clinic to another clinic," she explained. "And the third part of the datasets is the history of the patients themselves. Some of the patients, we are noticing that they are having a high rate for missing their appointments and like the other patients. So that kind of history gives us an insight about those kinds of patients."

Importantly, for this project, "we did not address any kind of clinical data," she added, since that would require expert clinicians to decide which kind of clinical factors that might be affecting a no-show.

But using a basic dataset of patient information enabled creation of some initial models which were then validated to make sure which was best and most accurate.

"The project started two years ago. It takes phases in order to make sure that we are ready to [incorporate the model] within the electronic medical record system," said Al Ghamdi. "So in the first year the model was created, and I can say that we are in the stage of validating the model, this validation phase, it takes about four to six months.

"Part of that validation has been conducted within the data science, and then we launch it to a small group of clinicians and a staff from the nurses and patient services," she added. "And that phase, it took about another six months. At that point, it's been a year that we are validating and making sure that the model is reliable and we can really rely on the results from that model."

Once its data science experts were satisfied with the algorithm, MNGHA took the step to incorporate the model into its EHR system and integrate it into clinical workflows.

"The clinician can see that that patient scheduled for that day has a potential to not attend the appointment. And by having this kind of a flag within the medical record system, the clinician can send additional reminders, or, for example, asking the patient services to do a kind of call in order to remind the patient," said Al Ghamdi.

Eventually, the plan is to implements the model across all MNGHA facilities, in all regions.

For those health systems looking to try something similar for their own organizations, Al Ghamdi offers a bit of advice.

"Even if they start with a small dataset, it's better to do this kind of implementation, even in a small scoop of data or small list of parameters, because we know for sure that data is telling us a lot about our patients, and there are a kind of hidden patterns that we can discover by using the technique of machine learning and artificial intelligence.

"Taking the steps forward to deal with the data and gaining the knowledge out of it, it's something very important," she said. "It's a very simple model that can be created. But it has a huge impact on the organization."

Read a more in-depth case study about MNGHA's use of machine learning for predictive analytics here.

Mike Miliard is executive editor of Healthcare IT News
Email the writer: mike.miliard@himssmedia.com
Healthcare IT News is a HIMSS publication.