By Leslie Kernisan, MD

Should I be prescribing apps, and if so, which ones?

I recently came across this video of Happtique’s CEO Ben Chodor describing his company to Health 2.0’s Matthew Holt. In it, the CEO explains that Happtique is creating a safe and organized space, to make it easy for doctors to prescribe apps and otherwise “engage with patients.”

Because, he says “we believe that the day is going to come that doctors, and care managers, are going to prescribe apps. It’s going to be part of going to the doctor. He’s going to prescribe you Lipitor, and he’s going to give you a cholesterol adherence app.”

He goes on to say that they have a special process to make sure apps are “safe” and says this could be like the good housekeeping seal of approval for apps.

Hmm. I have to admit that I really can’t imagine myself ever prescribing a “cholesterol adherence” app. (More on why below; also found myself wondering what it exactly meant for Happtique to say an app was safe. What would an unsafe cholesterol app look like?)

Happtique, of course, is not alone in hoping that clinicians will be prescribing apps to patients in the near future. Many digital health enthusiasts expect that apps will become a routinely used tool in healthcare, especially if clinicians encourage patients to use them.

But under what circumstances will clinicians want to prescribe apps, and what would make it easier for them to do so?

Right now, my best guess would be that clinicians will mainly be motivated to prescribe apps in the same way that they’ve historically prescribed drugs:

• Because they’ll get marketed to, mainly by companies with financial interests in clinicians prescribing certain apps. This leads to clinicians both having an awareness of the app and having a feeling that prescribing the app is a good thing at some level (whether or not this is true by other objective evaluations).

• Because someone will make it very easy for them to recommend an app to patients. This could be a platform such as Happtique becoming dominant enough – and usable enough – such that it becomes very easy for a clinician to prescribe an app, just as Amazon has made it very easy to buy on their site. Or it could be a major medical system integrating links to a smaller curated set of apps into their e-prescribing or patient recommendation system. (Happtique is working on this, too.) There are lots of ways to make it fairly easy for clinicians to take certain actions, especially if someone stands to make money as a result.

• Because patients and families might request an app. Just as direct-to-consumer marketing of drugs can affect prescribing, direct-to-consumer marketing of apps could influence clinicians. This could be a general request: “Can you recommend an app to help me with this health problem?” Or it might be a request that a clinician endorse a specific app: “I’ve seen ads for a Beer’s Criteria app. Would you recommend I use it, to help me spot medications that could potentially harm older adults?” (In which case I’d likely answer a resounding YES; a geriatrician can dream, no?)

Ok, now let’s step back a bit and think about what the above approach has overall brought us when used by the pharma industry:

• Lots of prescribing of drugs, whether or not drugs were the optimal approach to managing a given problem
• Lots of prescribing of certain well-marketed drugs, whether or not those specific drugs were the likely to be the best choice based on available evidence
• Lots of profit for certain companies
• High pill burden for many patients
• Substantial medication-related expenses, both out-of-pocket and for payers
• Harm related to medication side-effects and interactions

Hmm. Obviously many have also benefited from the prescription of pharmaceuticals, but still, especially when one considers the med lists of elderly patients, it becomes clear that there’ve been downsides to the way clinicians have been historically been encouraged and enabled to prescribe.

Will we do better when it comes to apps?

Why should a clinician recommend an app?

If I were asked “Why should a clinician prescribe an app?” I would answer as follows:

Because it’s likely to help the patient reach his or her most important health goals, and is a good fit within an over-arching medical management plan.

In other words, if the goal is to provide sensible medical assistance to patients and families, the use of an app should be likely to:

• Help a patient work towards the most important medical goals.
o This means clinician and patient should’ve discussed goals overall, and prioritized which issues are most important for the time being. Since I take care of complex older patients, prioritizing issues is really a must, and then we can set certain goals for the issues we’ve decided to focus on.

• Be likely to provide benefit or otherwise be clinically useful.
o This doesn’t mean we always need peer-reviewed studies demonstrating that use of this particular app provided a health benefit. But there should be some reason to believe using an app will be clinically useful.
 This could be because the app facilitates collection of data needed to revise the treatment plan, i.e. documents pain, incontinence, sleep patterns, as-needed medication use, etc.
 Or it could be that the app digitally guides patients through an intervention previously found to be beneficial, such as a home exercise plan.
o As with the prescription of a drug, recommending an app should include guidance as to what benefit the patient can expect, as well as a plan for ensuring that the app is delivering benefit as expected.

• Be a good, feasible fit within an overall management plan.
o Just as I don’t prescribe a medication in isolation, without considering the patient’s other medical conditions and other prescriptions, I wouldn’t recommend an app in isolation.
o I find that most patients and families have only so much bandwidth available for daily healthcare management tasks. So in considering an app I’d also try to be mindful of how many other apps have been recommended, and I’d try to work out an overall plan that was going to be manageable for the patient. After all, there is only so much futzing with devices that one can do in a given day.

Since my patients are older adults with multiple medical problems, I expect that I wouldn’t very often suggest apps that are narrowly focused on something like cholesterol. I don’t need patients to “adhere” (a problematic word for many reasons) to their statin and learn all about which diet is best for lowering cholesterol.

What I do need is for patients to be supported in taking several meds that we’ve decided on, and then I need them perhaps to have support in remembering whatever combination of diet tips we decided was a reasonable fit for their preferences and combination of medical conditions. (For example, in some cases I *do* advise the family of a frail elderly diabetic to loosen up and let the patient have a doughnut.)

Also, it would be burdensome if every specialist my patient saw decided to prescribe their own pet app for “adherence” to whatever condition the specialist was concerned about. Just as Boyd et al demonstrated in their 2005 JAMA paper that attempting to implement all guideline-recommended care for nine commonly co-existing chronic conditions led to an unmanageable plan of care, prescribing an app for every little thing on an older patient’s problem list will definitely lead to app overload for the patient’s care circle.

In short, I can envision apps helping patients and families manage a medical care plan. But I worry that we’ll end up making the same mistakes with apps as we’ve often made with the prescription of medications: recommendations based on marketing rather than thoughtful assessment of expected value, and prescription of apps for every little medical condition rather than choosing a few high-yield apps based on a whole-person approach to managing healthcare.

Ensuring thoughtful clinical app use

How to ensure that the clinical recommendation of apps is thoughtful and person-centered? I’m not sure, but in general I think there would be value to clinicians and patients doing the following:

• Review use of the app in the context of the overall big picture of the person’s health, and the overall goals of medical care
• Be explicit about the purpose of the app and expected benefit
• Plan a future time to review use of the app and assess whether the benefit justifies continued use.
• Periodically consider winnowing down the number of apps being used, especially if the patient or care circle report any app fatigue.

You might notice that the above looks an awful lot like what we should be doing – but often don’t – with patients’ chronic medication lists.

Summing it up

Apps, like pharmaceuticals, can in principle help patients and families meet their healthcare goals. Many would like clinicians to embrace apps and begin recommending them to their patients.

It would be easy for clinicians to end up making the same mistakes with apps as we’ve often made with the prescription of medications: recommendations based on marketing rather than more considered assessments of expected value, and prescription of apps for every little medical condition rather than choosing a few high-yield apps based on a whole-person approach to managing healthcare.

To ensure more thoughtful recommending of apps, especially for medically complex patients, we could consider strategies that can be helpful in managing multiple medications. These include reviewing the use of a proposed app within the context of the patient’s overall health issues and goals of care, being explicit about the purpose of the app and expected benefit, and periodically reviewing and adjusting app use. The recommendation of apps for every single medical diagnosis affecting an older person could easily lead to app overload, and should be avoided.

Leslie Kernisan, MD, MPH, has been practicing geriatrics since 2006, and is board-certified in Internal Medicine and in Geriatric Medicine. She blogs at GeriTech.