I was helping a resident prepare to place a needle into a patient’s knee joint. The patient had come to the hospital because of excruciating pain that was much worse than what he experienced from his chronic arthritis, and perhaps it was the sight of the needle that prompted him to ask me, “Has a patient ever hit you?” This startling question made me pause for a moment. I remembered that, in fact, I had been hit several times. Once when looking into the ear of a woman who I had thought was comatose; she landed a hard right to my cheek, knocking me to the floor. Another time, an intoxicated man kicked me as I was attempting to check his legs for fractures. And then there was the patient who pulled a gun (fortunately it was a starter’s pistol, but what did I know?) as he demanded to know the results of his wife’s laboratory tests. I described these incidents without much emotion, much as I might have described previous minor traffic accidents, while the resident cleaned the patient’s knee with antiseptic solution and placed a surgical drape over it. As the resident unwrapped the syringe I asked our patient why he had asked me that question.
“Just curious,” he said. “And I noticed that needle. Do I really need it? Is it going to hurt?”
“We will numb you up. I hope that will take away any pain, and once we remove fluid from your knee, you may actually feel better,” I said. “Without examining the fluid we would not know whether the pain is being caused by an infection that might need antibiotics.”
I did not know how much to explain our investigation of the literature on how to best approach his problem. The resident and I had looked up a systematic review on septic arthritis by Mathews et al1 that said: “Little good-quality evidence exists to guide the diagnosis and management of septic arthritis. Overall, no investigation is more reliable in the diagnosis of septic arthritis than the opinion of an experienced doctor.” I had weighed various elements of the history (the pain had come on suddenly and was severe; he had felt hot; he had a history of pseudo gout). I had also considered the physical exam (no fever, some joint effusion). The white cell count was slightly elevated. And then I had weighed the risk of missing the diagnosis of a septic joint (possible destruction of the joint, sepsis, death) versus the risk of sticking a needle into an uninfected joint (pain, possible introduction of infection). So here I was, the experienced doctor who was not sure. After discussing options with the patient, he encouraged us to stick a needle into the joint and find the answer.
“Well, I don’t want to hit you,” he said.
“That’s good,” I said.
“But at least it would not be the first time.”
“True. But I really don’t want it to happen again. I’m getting older. Those other times were long ago. Do you want to talk about this procedure some more and think about it? I think we have gone over all of the possible complications. But we don’t have to do this.”
“No, go ahead.”
“Okay, but please warn us if you are having any pain. Maybe we should give you something to help relax you, sort of make you sleepy.”
“No. That’s all right. I want you to do it. I want to be awake.”
I looked at the resident. I had explained that the resident was going to do the procedure, and I was really there to supervise and to assist in case he had trouble. But now the resident was becoming apprehensive. Since he had done this procedure only a few times, we had reviewed where to stick the needle and how to withdraw the fluid. After the resident numbed the skin, he began to insert the needle into the knee just under the patella. His hand shook, and he moved the needle forward so carefully and so slowly I could barely detect any forward motion. I was watching the patient for any reaction and was ready, just in case. “Any pain?” I asked.
The man grimaced.
“I think I’m hitting bone,” said the resident. I looked at the needle and the syringe. There was no sign of any fluid from the joint.
“Just come out and angle the needle down a bit,” I said. He tried to reposition the needle.
“Bone again,” he said.
“Okay, let me see what I can do,” I said. I moved over to the other end of the bed and regloved. The resident gave me the syringe and needle. I looked over at the patient. His eyes were open wide, glaring at me, and his arms were flexed with his fists against his cheeks. “All right, sir,” I said. “Another little poke.” The needle slipped under the patella and the joint fluid came out easily. “All done,” I said.
The man relaxed his arms and began to breathe easily. “What do you think, Doc?”
“I don’t know. We have to send the fluid to the lab. And we will look at it under the microscope. But it looks fine, normal to me. I hope it didn’t hurt too much.”
“No, it didn’t hurt at all. It was just the anticipation of it. But when I thought about hitting you, it wasn’t so bad. Do you know what I mean?”
“Not really,” I said.
“Well, I figured if you were willing to take the risk that I might hit you, and still do this, it must be important. It was like we were in it together.”
I nodded and went off to look at the fluid with the resident. Fortunately, there were very few white cells and no crystals. It was probably just osteoarthritis, which we would be able to treat with anti-inflammatory medications. Later, the cell count returned from the laboratory confirming our impression, and we were able to discharge the patient with the appropriate medications.
As we discussed the case, the resident said: “I am sorry. I guess I was nervous. All that talk about hitting you.”
“Yes, that was strange. But, then again, we were standing there with a sharp needle pointed at his knee, which already hurt. I am glad he did not have a septic joint, but I feel bad that we had to stick the needle into his knee to prove it. Unfortunately, the literature on this issue is not clear. Even the guidelines that we found recommended examining the joint fluid. Unfortunately, guidelines are only as good as the research, and as we found out, the research is not good. And in any case, we always need to put these guidelines into the context of our own patient.”
That discussion reminds me of a group of articles we have in this month’s journal about clinical care guidelines.2–4 I first remember clinical care guidelines being promoted as a scientific alternative to physician clinical judgment, which could vary from physician to physician. Guidelines typically have originated with a question, such as “How should a physician best diagnose and treat a swollen painful joint?” Experts have then convened to review the evidence, conduct a systematic review, and issue recommendations such as when to remove fluid from the joint (as we did), how to make sense of the results, and what type of treatment to offer. In 2011, the Institute of Medicine published a framework for clinical practice guidelines5 that presents the key elements that should be used in formulating a guideline.
In their excellent review of the history and current use of guidelines in this issue, Sox and Stewart2 identified six elements that should be included to incorporate guidelines into the process of continued learning and process improvement:
- A systematic review of the pertinent evidence
- Recommendations for action that specify the necessary steps to ensure consistent execution
- Representation of the standard of practice in a form—such as an algorithm flow chart—that facilitates testing of each decision point
- Using the clinical standard in daily practice while documenting patient data at each decision point
- Encouraging clinicians to comment or take alternate actions but insisting on explanations
- Rigorous analysis of the data to advance and refine the care process
Sox and Stewart emphasize the need for ongoing assessment and refinement of a guideline based on clinical practice. This is precisely what Farias et al3 attempt to do in their development of standardized clinical assessment and management plans. These are similar to clinical guidelines but provide the added feature of an iterative process to encourage providers to diverge from a guideline when appropriate, explain their reasoning, and then use that information to adjust the guideline for the future. It would appear to me that for such a system to actually improve guidelines, providers would have to be very careful to diverge from guidelines only in cases where their approach had strong evidence to support it. I imagined that one would have to be very experienced, careful, and confident to behave outside of established guidelines.
Mercuri et al4 investigated this issue by documenting differences in adherence to clinical practice guidelines by experts and novices. They presented cases to 28 expert emergency physicians (residency trained, at least five years of experience) and 28 novice emergency physicians (residents in the first two years of emergency medicine training) and found that the experts deviated from the clinical guidelines more frequently than did the novices when presented with information, such as patient occupation or rural residence, that could increase the risk of and affect the consequences of following a clinical guideline. This study points out the problems of judging a provider’s quality of care based on adherence to a clinical guideline, because experts will take into account the specific contextual factors and deviate from a guideline if doing so appears to be the best choice for the patient. Less experienced providers may not be willing to take that risk. This study also lends credence to the approach of Farias et al3 to collect data that might help modify and inform clinical practice guidelines based on the nonadherence actions of expert physicians.
These studies suggest that as we educate our students about the value of evidence-based medicine and the use of clinical guidelines to inform our diagnostic and treatment decisions, we should also remind them of the importance of adjusting guidelines to fit the circumstances of the specific patient. Only through efforts to get to know our patients as unique people will we be able to elicit the social factors, personal experiences, and individual values that should shape the application of guidelines. We have to establish the kind of rapport that helps our patients believe that “we are in it together” and that they can trust that our advice represents an understanding of their unique concerns rather than a blind adherence to a standard guideline.
We sometimes forget that we may need to invade the physical and emotional spaces of our patients in our efforts to establish rapport, discover and treat their problems, and provide effective advocacy for their ongoing needs. Although a patient’s reaction of striking out at a doctor or nurse, physically or verbally, may seem indefensible, if we imagine it as a manifestation of the patient fighting for his or her autonomy in an unfamiliar and sometimes painful and frightening environment, it may be more understandable. Many of our patients inhabit a world in which they experience the destructive effects of imbalances of wealth, power, and social status every day, and in which mistrust is a survival mechanism. Why should I think my white coat would change that? And so the next time a patient asks me, “Has a patient ever hit you?,” I hope that, rather than reeling off past incidents like notches on a belt, I will take a step back and consider what led to the question. And then I will ask the patient what he is afraid of, and how I can help.
David P. Sklar, MD