When we were mulling over a title for our forthcoming book on contextualizing care, our friends begged us not to use the words “context” or “contextualizing” in the title because they thought the terms sound boring. We appreciate where they are coming from, but it’s also hard to reconcile with what we’ve come to appreciate over the years: attending to patient context when care planning is the one and perhaps only essential element of quality that we still need doctors for. Computers are more accurate and less costly at diagnosing and treating medical conditions. What they can’t do is get to know a person and figure out what they really need.
Every person’s life is infinitely complex. One life, at any point in time, is an amalgamation of relationships, financial issues, responsibilities, dependencies, capabilities, blind spots and motivators plus the ever present effect of chance. When a person stops taking their meds, or loses control of their asthma, or has to decide whether to put off surgery to care for a sick child, the possibilities are myriad and at some level fundamentally unique. No two people ever stopped taking their meds under exactly the same circumstances. One individual might stop because they have transiently lost health insurance, and free samples of the medication will carry them over. For the other who has no prospects for coverage, finding an alternative less costly and perhaps less effective therapy might be the only option. Planning care for a poorly insured patient with Hepatitis C for which medication can be costly and coverage sparse, requires knowledge of the patient’s resources and options.
Contextualizing care isn’t just helping a patient solve a practical problem. For instance, a woman is devastated when she learns that her spouse is abandoning her for someone else in her hour of need, after she is diagnosed with cancer. Her sense of self is crushed and she acknowledges feeling hopeless and ready to “give up.” Her physician recognizes the context, which is that this woman feels terribly alone and needs more than a sanitized response. Rather than reacting with just sympathy, her physician shares with her the anger and indignation he feels on her behalf, saying “Are you going to throw in the towel so that jerk who once professed to you his undying love and commitment can walk off with everything you own and give it to that woman he’s moved in with? What you need is the nastiest divorce lawyer in town….and I’ll help you find that person. Make him pay and see how long his girlfriend sticks around!” A smile comes over her lips and a true healer-patient alliance is formed.
Thinking contextually is about identifying what in the infinite complexity of every moment is most relevant to a particular situation. Everyone does it. My colleague, Simon Auster who for 30 years taught a course titled “The Human Context in Health Care” once put it this way: You’re driving home late from work because of an unexpected delay, concerned because it’s your wedding anniversary and you and your spouse have reservations at a fine restaurant that you’ve both been looking forward to for some time. The speed limit on the highway is 50 miles an hour but you feel safe driving faster. At the same time, you know the cops pull cars over on the road if their going over 55. However you haven’t seen cops at this hour and you don’t think you’ll make it in time unless you speed up. That said, it rained earlier and the road is still wet…balancing all these factors you drive at 60 miles an hour. That set point is the product of contextual thinking.
Thinking contextually can be a disaster when you don’t know and can’t manage yourself well. For instance, tending towards a sense of infallibility could prompt you to drive at 90 miles an hour under the aforementioned conditions and cause a terrible crash. In the clinical context, showing one’s anger in a patient interaction could be destructive instead of constructive if it was unleashed because of the clinician’s unresolved personal issues rather than as a stimulus to promote self-empowerment and forge a bond.
We’ve developed methods for assessing attention to context in care planning, based on audio recording clinical encounters. We’ve documented considerable variation using this quality measure, and shown that when a care plan addresses contextual factors patients have better outcomes at lower costs. Overall clinician performance at contextualizing care, however, is disappointingly low. In a study employing an experimental design, only 22% of encounters demonstrated attention to context. In that same study, however, physicians perfectly followed guidelines about 73% of the time. If a patient’s asthma was getting worse because he could not afford his medication and wasn’t taking it, physicians were more likely to add additional medication intended for treating worsening asthma than address the access problem by, for instance, switching the patient to a less costly generic. That’s treating a disease rather than the patient with a disease.
73% isn’t bad but it’s not as good as a computer if the goal is algorithmic care. And often that’s not the goal. When physicians are only good at treating disease but not people, they are on a road to becoming obsolete and patients aren’t getting what they need.