How Great User Experience (UX) Might Be Killing You
If you read my newsletter, you are aware that the past two years have been a bit challenging for me in some respects. I have gone from being an avid long distance runner to now walking with a limp. My care up until recently was with one of the biggest integrated managed care consortiums in the United States. Unfortunately, in my case, working with them has been a disaster, and I am now shopping for a better way to treat my condition. However, that is not what this post is about. I am sharing it because my personal trials and tribulations treating a painful hip created the impetus for exploring a direr situation — a situation on which I believe my eclectic professional experience with UX and healthcare gives me a unique point of view.
There are pros and cons, strengthens and weak points, within any complex system — the healthcare system is no exception. When weak points are not looked at critically, that is when we can run into serious problems.
A National Emergency
In 2013, a group of scientists from the Johns Hopkins Bloomberg School of Public Health, Johns Hopkins University, Stanford University and Mayo Foundation for Medical Education and Research provided hard evidence that something was not right with the opioid prescription rates in the United States. When they analyzed the prescription rates of opioid and non-opioid therapy for non-cancer pain in the period between 2000 and 2010, their data revealed some interesting trends. Although patients’ reports of pain remained unchanged in this period (pain was the main cause for one fifth of all visits to a doctor), opioid use nearly doubled (it went from 11.3 percent to 19.6 percent). Just as intriguing was their finding that the use of non-opioid pharmaceuticals did not follow the same trend. Only opioids were on the rise. Researchers struggled to find any association with patient, physician or practice characteristics that could explain the increase (Daubresse et al., 2013).
Leonard Paulozzi and his colleagues from the Centers for Disease Control and Prevention (CDC) cite even gloomier figures. They report that the sale of prescription opioid drugs keeps increasing; in fact, they quadrupled in the period between 1999 and 2010. In 2010, enough opioids were prescribed to medicate every adult American for a month (Paulozzi, Jones, Mack, & Rudd, 2011). To the best of my knowledge, this data has not improved in the past couple of years. The issue became so widely recognized that in 2014, a White House Summit was held to address the impact of opioids on American society (Clarke, Skoufalos, & Scranton, 2016), and just a couple weeks ago President Trump acknowledged that the opioid crisis is a national emergency.
Furthermore, the opioid epidemic has led to a significant increase in addictions, drug overdoses and deaths among Americans. To put this in perspective, the number of deaths from drug overdoses is now approaching the number of road deaths per annum. More people are dying of opioid poisoning than “street drugs,” drugs like cocaine and heroin (Paulozzi, Budnitz,& Xi, 2006). Many health experts are warning that this epidemic is worsening. The current state of affairs is having a grave impact on public health and safety, not to mention the financial burden it is putting on health insurance (estimated at $72.5 billion per year as reported by the National Drug Intelligence Centre).
Many health professionals offer strategies on how to better manage pain and minimize the risks of opioid overuse and misuse. However, we should not forget the fact that the number of patients experiencing pain has remained about the same over the past decade. Nonetheless, opioid prescriptions have been increasing. Why is that?
Hard Pill to Swallow: Is Patient-centered Design Part of the Problem?
I have built a reputation for my expertise on user experience and healthcare. A significant portion of my working hours is spent innovating ways to improve the satisfaction, engagement and fun end users have with digital health products and various fitness consumption modalities. While getting my Ph.D., I also did a lot of practicum work assisting physicians with employee burnout protocols (through an affiliation with a hospital group in the Bay Area; important to note, it is not the same group mentioned at the beginning of this post). During my work as a doctoral candidate on physician burnout, after malpractice lawsuits, the number one thing I heard that contributed to physician burnout was the cognitive burden a company called Press Ganey put on this particular physician group. If you are not familiar with Press Ganey, they are one of largest companies in the “patient satisfaction” business and their purported “mission” is to “support health care providers in understanding and improving the entire patient experience.” Simply put, they collect patient feedback data on doctors, and many times these assessments are tied to a physician’s livelihood.
More and more doctors and health experts have been recognizing that the increase in opioid use is correlating with the increasing emphasis on patient satisfaction. This might seem paradoxical at first (it did to me; after all, design thinking and great UX is about empathy and making the end user happy, right?), but it actually makes a lot of sense once you look at the facts at a macro level. Jenice Clark and Alexis Skoufalos of Thomas Jefferson University, Philadephia, and Dr. Richard Scranton of Pacira Pharmaceuticals conclude their article on the opioid epidemic in America with an insightful opinion from an expert panel:
It takes mere seconds for a primary care physician to write a prescription for an opiate; it takes 30 minutes to explain why he/she is unwilling to do so. (Clark, Skoufalos, & Scranton, 2016, p. S-7)
After digging further, I found patient satisfaction data now also gets reported via the HCAHPS survey. This is becoming a widely used healthcare quality metric. In fact, it is becoming the national standard for collecting the “patient experience” — so much so, scores from these surveys are starting to be incorporated into calculating reimbursements. In other words, in these cases, happy, drugged up patients indirectly mean more resources for a physician’s practice. Although patient satisfaction is a very subjective metric, it is used quantitatively to evaluate and compare physicians and determine financial compensation, job retention and promotions.
Are There Cases Where Grading UX Should Not Be Measured by the User?
Some experts warn that the “positive patient experience” is not necessarily linked to better health outcomes either. For example, in an article titled The Cost of Satisfaction, Joshua Fenton of University of California, Davis and his colleagues found that patient satisfaction was linked to:
- higher admission rates,
- greater overall expenditure,
- higher prescription drug use and
- increased mortality rates (Fenton, Jerant, Bertakis, & Franks, 2012).
Their nationally representative sample showed that satisfied patients did not necessarily fare better when compared to those that left their doctor’s office less satisfied. The rub: they are actually more likely to die!
Furthermore, studies show that patients often ask for services that are not clinically indicated. For instance, Kravitz et al. (2005) found that patients’ requests have a big effect on physicians prescribing antidepressants. Also, physicians whose patient rating affects their livelihood are more likely to budge and agree to patient requests (Pham et al., 2009). Patients expect their demands to be met. If unsatisfied, patients often now have a direct lever to affect a physician’s livelihood.
I Would Like the Credit, but I am Not the First to Stumble Upon This
Aleksandra Zgierska, Michael Miller and David Rabago of the University of Wisconsin-Madison, Madison (2012) believe that patient satisfaction surveys are generally a helpful initiative and can be a driver of positive change. However, they can also contribute to the prescription of addictive medications such as opioids.
When looking at these “patient satisfaction” instruments critically, you will often find the first question on the survey is an enquiry if you were satisfied with the way your doctor treated your pain. Since modern day primary care physicians experience time pressures, time-consuming discussions on alternatives to opioids are not necessarily promoted (for more on the burdens of modern healthcare, see my interview with Matthew Heineman about healthcare in America). When faced with high patient volumes and demands for interventions, it can be easier for doctors to simply write a prescription. In this way, the patient often leaves the (brief) office visit initially happy.
After working with burnt out physicians for several years now, I have an immense amount of empathy for this employee group. As such, it is very important to note that the harm here is not just to patients. As Zgierska, Miller and Rabago warn, these practices also can leave clinicians in emotional and moral distress. In fact, it is likely that the opioid epidemic is a silent yet significant factor in the explosion of cases of physician burnout. When you peel it back, it really is becoming a no-win situation for everyone.
When Great UX Gets in the Way of Great Outcomes
In my case, I likely need a hip replacement, but the integrated managed care consortium I was a part of does not like to perform replacements on people under 50 — especially males. Why? Although younger candidates have better surgical outcomes, this group (my strata) also has higher readmission outcomes. Having to get a second hip replacement is a more complex procedure and often has much graver outcomes than an initial replacement.
The most egregious aspect for me was that even though this consortium was impeding my effort to get better, the medical staff would still ask about my activity level during every office visit (by way of the worn out script of their rudimentary health risk assessment, affirming to me there was no real understanding of my condition). I’d have to inform them every time I was there because I would love to be more active. Instead of ever really developing a path forward, I’d routinely leave with a prescription for NSAIDs and/or opioids. It was more advantageous for them to keep their costs down and alleviate my symptoms than address the problem.
When I started to complain to other colleagues in healthcare, I quickly realized this is the elephant in the room. The over-prescribing of opioids is additionally concerning when many suggest that long-term use for chronic, non-malignant pain has more cons than pros. It brings many side effects, including gastrointestinal issues, confusion, respiratory problems and increased risk of infections and tumor growth (Clarke, Skoufalos, & Scranton, 2016). Inappropriate prescribing also means that more drugs become available to the general public through black markets, leading to further abuse and addictions. Ultimately, I have now stopped taking all prescribed painkillers and just suffer through the pain until I can treat the condition and not the symptoms. My current UX is terrible but at least I’m not a dope.
Do You Want Great UX, Or Do You Want Great Healthcare?
When it comes to improving the quality of care in our hospitals, I am not attempting to marginalize that pain assessment and pain management are two important standards. However, patient satisfaction — originally designed to promote quality of care — has in some cases undermined the principles of good medicine. This may sound contrarian from someone who generally advocates for patients to become the smartest person in the room about their particular condition, but maybe some decisions should be left to the people most qualified to make them. We need to foster systems that afford doctors the luxury of treating conditions and not simply addressing symptoms.
Instant gratification has its place, and it is natural to want to avoid discomfort. However, in this particular narrative, it is not a good replacement for legitimate treatment options. The current system is rigged to make you sicker. That has got to change, especially in cases like mine where there are better treatment options. Great UX is not always in the user’s best interest. In fact, as this post outlines, it can lead to many ill effects — on a personal, national and global level.
Sources & further reading:
Clarke, J. L., Skoufalos, A., & Scranton, R. (2016). The American Opioid Epidemic: Population Health Implications and Potential Solutions. Report from the National Stakeholder Panel. Population Health Management, 19 (Suppl 1), S1-S10. doi:10.1089/pop.2015.0144
Daubresse, M., Viswanathan, S., Alexander, G., Yu, Y., Chang, H., Shah, N., & … Kruszewski, S. (2013). Ambulatory diagnosis and treatment of nonmalignant pain in the United States, 2000-2010. Medical Care, 51(10), 870-878. doi:10.1097/MLR.0b013e3182a95d86
Fenton, J., Jerant, A., Bertakis, K., & Franks, P. (2012). The Cost of Satisfaction A National Study of Patient Satisfaction, Health Care Utilization, Expenditures, and Mortality. Archives of Internal Medicine, 172(5), 405-411.
Kravitz, R., Franz, C., Azari, R., Wilkes, M., Hinton, L., Franks, P., & … Feldman, M. (2005). Influence of patients’ requests for direct-to-consumer advertised antidepressants: A randomized controlled trial. Journal of the American Medical Association, 293(16), 1995-2002. doi:10.1001/jama.293.16.1995
Paulozzi, L., Budnitz, D., & Xi, Y. (2006). Increasing deaths from opioid analgesics in the United States. Pharmacoepidemiology and Drug Safety, 15(9), 618-627. doi:10.1002/pds.1276
Paulozzi, L., Jones, C., Mack, K., & Rudd, R. (2011). Vital signs: Overdoses of prescription opioid pain relievers — United States, 1999–2008. Morbidity and Mortality Weekly Report, 60(43), 1487-1492.
Pham, H., Reschovsky, J., Landon, B., Wu, B. & Schrag, D. (2009). Rapidity and modality of imaging for acute low back pain in elderly patients. Archives of Internal Medicine, 169(10), 972-981. doi:10.1001/archinternmed.2009.78
Zgierska, A., Rabago, D., & Miller, M. (2012). Patient satisfaction, prescription drug abuse, and potential unintended consequences. JAMA – Journal of the American Medical Association, 307(13), 1377-1378. doi:10.1001/jama.2012.419