Open enrollment starts this week and millions of people will receive reminders from their employers and carriers to consider coverage changes. This year many will factor in their recent or expected encounter with the Coronavirus. It is a solemn reminder that planning ahead not only protects your health but also your wallet.
But what happens if you forget to plan?
When we start our day we unconsciously run into our prior day’s or month’s plan. The alarm clock on your smart phone goes off reminding you to get ready for work and the dog licks your foot to get up and feed him. Other reminders are things like Google
Before the Industrial Revolution and the requirement to show up for work for a given time period and clock-in to operate a machine, people could arrange their life more flexibly as they were rewarded by output, not time served. The nine-to-five existence we all know has really only been around since around the middle of the 19th century and made constant reminding necessary.
Reminders are a great thing in this increasingly busy world. Mortgage pay reminders reduce risk of a 60-day delinquency by 21% boosting your credit score by 10 points. Reminders also have negative effects on our cognitive wellbeing because they distract us mid-work. Shriya Sekhsaria, CEO of The Memory Jar Company, shared with me that daily interactions went up by 4 per user when simple daily reminders for personal photos of life events are ready to be put in their digital time capsule. This actively combats the organic drop off such services see when users initially sign up plus it also generates more ad revenue for the company. A recent article by Danish researcher, Christina Gravert, illustrated the effect of properly structured reminder cadences reducing opt-outs. She recommends to use reminders not just based on demographics but consumer behavior. In a healthcare setting, Dr. Nesochi Okeke-Igbokwe says, “I may see a patient who informs me that they have diabetes. After I enter the patient’s diabetes diagnosis into their electronic health record, then I will get CDS (Clinical Decision Support System) notifications at the top of my screen reminding me that this patient is diabetic and they may be due for their annual eye exam, diabetic foot exam, or blood work to check their Hemoglobin A1C level. Such reminders are truly valuable tools, and when implemented effectively will ultimately translate into improvement in the quality of care a patient receives.”
Such medical treatment reminders have a verified positive effect on outcomes. A recent NIH article documented that reminders result in 25-40% of appointments being rescheduled or cancelled allowing for reallocation of time to other patients. Patients without reminders rescheduled only 8-12% of appointments. This has a direct effect on healthcare cost, delay of treatment, unnecessary hospital admissions and adverse effects. Therefore, reminders are a very inexpensive way to remedy skyrocketing healthcare cost and suboptimal outcomes.
The study also highlighted the major challenges that remain, such as inaccurate patient records around preferred language and cognitive ability, incorrect contact information and lack of confirmation feedback, improper timing of contact and interaction channel availability, and an overall deficit in prioritizing context around high-risk groups . Zendesk’s study on customer experience judges “traditional reminders” harshly. More than half of Millennials and a similar portion of older generations find “unsophisticated” reminders; meaning repeat, canned outreach based on a simple date/time setting without additional information and context, intolerable. This goes in line with the Danish research mentioned earlier.
This flexibility and variability of best actions on a large scale is a problem for many businesses who operate based on timing of your decisions and actions. So why do many, if not most of them, not support the correct timing of your actions, especially if it is in their commercial interest?
Leading organizations provide this support for similar high-value and/or common processes. Think how long it took for your wireless company to send you a reminder to activate your phone and start using it. They want you to ultimately consume more data and services as early as possible. Your credit card company wants you to start using your card the day you receive it in the mail. The doctor wants you to be on-time for your appointment so he can charge your health insurance carrier earlier and not sit idle for an hour. The delivery service wants you to receive a shipment so the driver does not have to return a second time.
All the required information is not only available internally to automatically remind you of an incoming shipment, plus it is happening on a set interval. All these scenarios also keep cost for such consumer services low because they reduce uncertainty.
The real challenge with something so unexciting as a reminder is when it requires multiple parties and more complex decision criteria to trigger the infamous next-best-action (NBA).
Over the years, I have run into fairly complex NBA examples. One of the most interesting ones was at a consumer electronics retailer. This organization required a team of mathematicians, data scientists, developers and strategy consultants to come up with a technology and process framework on how to handle product returns. The resolution path of a return basically boiled down to a few criteria to establish the total economic value of a transaction. The criteria were the customer life-time-value, the item’s purchase price and book value, the availability of like substitutes and the remediation cost (replace, resell, scrap or repair). Again, though, the data pools to calculate this value were all available in-house, including product catalog, vendor cost estimates and customer transaction history.
Timing Is Everything
While timing was not a paramount issue in this example, other examples in consumer products, automotive, insurance and banking, certainly were. In their world, timing is everything because it dictates extended warranty upsell, a trade-in, premium revenue, loss ratios, regulatory adherence and customer retention.
Case and point; a regional retail bank, which should have reminded its customers, but to their Chief Strategy Officer’s chagrin didn’t, that their recent and unusually high deposit amount from a real estate sale, for example, should be used to finance a new house or be invested through the bank. The CSO knew that they had a 72-hour window to make a compelling proposal to these customers, otherwise, the funds would be transferred and invested with another financial institution. A simple reminder to a relationship manager would have been a step forward but ideally; a best mode-of-contact, timing, and contextually relevant message or offer would have been ideal. More importantly, additional discussion paths, content and reminders-to-act should have been part of the customer experience.
One would also believe that a consumer goods manufacturer certainly should seek to work through its retail partners or smart devices to obtain some sort of product activation signal. With increasingly more instrumentation in home appliances and direct-to-consumer (DTC) purchases, the eCommerce application or equipment itself should ask for a warranty registration within a certain timeframe. In this day and age, relying on a consumer to fill out a paper warranty registration card is just plain crazy. Companies may believe that it saves them from a product warranty repair expense but they leave hefty margins on the table. In addition, this menial task may actually present another touchpoint opportunity to educate the consumer or upsell consumables like filters or attachments.
Timing is also everything in healthcare. Find out you are sick too late, forget to pick up your prescription, don’t check in with your support group or doctor, and risk to your health increases. The associated long-term cost spirals out of control for the carrier, which passes it back to you in form of premium hikes.
The challenge around timing stems from the network effect required to reduce risk. If every party were to share the same system to capture, analyze and trigger a reminder – see the CDS example I mentioned earlier – or respective systems were integrated, we would not have any issue. The reality of integration, centralization and data quality looks much different.
Healthcare providers mainly interact with insurance companies around patient matters when service authorization or invoicing happen so not all patient interactions are within the carrier’s observable universe. Even doctor’s may only receive a prescription pick-up alert if the issued an e-prescription. Support groups are often not associated with carriers at all, as services beyond counseling are typically not covered. On top of this, data privacy under HIPAA is also a major complexity factor.
Despite the vested interest of the insurance carrier to know that a patient follows a best-practice treatment path from research online, initial diagnosis, follow-up doctor’s visits, lab work, prescription regimen to communal support meetings; it is difficult for the carrier to collect all of these and detect deviations. This puts the patient at risk and the carrier at a higher loss risk. While patients are ultimately responsible for their health, we cannot assume they have the expertise and tools to do so. More importantly, key opinion leaders like Dr. Vartabedian, indicate that even physician’s continue to struggle with integrated, big data-driven decision making. Interactions are spanning weeks and months, can occur on the carrier’s general or patient care portal, are often only traceable through claims and involve multiple outside parties and highly sensitive information. Care paths are very complex and vary dramatically by factors like comorbidities, demographics, life style and even drug regimens and their interactions. Capturing this in software for repeatable and scalable use is difficult, costly and requires ongoing governance.
Pharmaceutical manufacturers are in an even worse position. Aside from any online research, advice and financial assistance they may offer patients for a drug regimen, they really only know what prescription was sold through what pharmacy network in a given zip code. For marketing purposes, they use data brokers to understand general prescription trends, which do not identify the patient. Pharma and medical device manufacturers can only try to back into which doctor prescribed a drug by gauging how many physicians of a certain specialty or with an event-based relationship would prescribe their drug, such as a statin, over a competitor’s product. This would then direct their physician, pharmacy and local market advertising spend. While information sharing concepts using a UPI (unique patient identifier) may have positive societal effects, I would suspect it to be used more for marketing rather than patient care in a private payer economy like the United States.
The Lack Of A Single And Complete Data Repository Shared Across All Healthcare Actors Puts Insurance Carriers In The Best Position To Orchestrate Actions
Ultimately, this lack of information centralization means that each party has an incomplete picture of the care journey. The insurance carrier may see signals on the early consumer research around an illness or physician based on their care hub application or an anonymous search on an affiliate website. The healthcare provider knows the current and historic diagnosis and treatment plan as well as prescriptions if they are all logged in the same application. The pharmacy or benefits manager knows about a patient’s coverage and prescription. Support groups know the diagnosis they work with.
In such networked scenarios it is incumbent upon the organization with the most directly managed consumer/patient profile and interaction data to collect the remaining “critical” journey/path signals to remind the individual or even an associated party, e.g. doctor, car repair shop, etc. that the next step is coming up or overdue. It is imperative to use the insurance carrier infrastructure from beginning to end of the care path as it collects the most touchpoints.
From a scalability perspective, the most pressing outreach could then be triaged by the carrier looking at high-risk individuals first based on related risk factors or time delay. DIY reminder apps, automated deliveries and sticky notes may be a solution but it addresses the care journey only partially. Using predictive models around care paths or customer journeys, we could then also predict which individuals are most likely to violate future next steps to mitigate even earlier or through different means, such as incentives. The point is that if a business can establish just a handful of optimal longitudenal paths, collect internal and external interactions and remind its customers, even without extra complexity in many scenarios, we may all be better off.