It is self-evident that less invasive surgery reduces trauma to the body’s tissues compared to conventional open chest surgery. Tissue damage is what triggers pain as patients recover from surgery, so it logically follows that less invasive techniques reduce pain. Published results about less invasive surgery in every field – cardiac, thoracic, general surgery, urology, GYN – conclude that postoperative pain is much better than open surgery. Pain evokes deep seated emotions (e.g. self-control and dependence on others) that make the promise of pain reduction a “selling point” for marketing expensive technologies like robotics that are used for less invasive surgery. Hospital websites that discuss tout their robotic programs specifically mention the benefit of less pain. This promise leads many patients to seek out surgeons that offer robotic surgery and pushes surgeons to go through the major effort needed to adopt these technically demanding methods.
Just about every aspect of robotic cardiac surgery is controversial, and the notion that smaller incisions actually reduce pain is no exception. I presented my initial 100 cases of robotic heart surgery at a national surgical meeting in 2008. At the time, I was asked by a well-known academic cardiac surgeon from Boston why pain was increased after robotic surgery in a way not seen with sternotomy. I muttered something about a learning curve for managing thoracic pain but didn’t have any data at the time and was too surprised by the question to provide a more thoughtful response. His question jolted me into realizing not everyone considered less invasiveness = less pain as a self-evident equation. It is hard to get patients and referring providers with little prior knowledge interested in novel surgical procedures without being honest about strengths and weaknesses. So, if I was going to continue using the party line that postop pain was reduced, I needed to find a better answer why.
Comparing pain levels after different surgical procedures can be tricky. A self-report system used clinically ranks the pain on a scale of 0 “no pain at all” to 10 “the worst pain possible.” Narcotics use is based on this scale and provides an additional metric of pain intensity. However, patients experience and remember post-operative pain in highly individualized ways. Some report extreme pain and demand large doses of pain medications while others seem to do well with much less medication. Patients with strong emotional well-being and support systems seem to perceive less pain and tolerate it more after surgery. Many patients don’t get adequate pain relief due to a variety of failures in staff and/or physicians to adequately assess and treat pain. All of this variability confounds the utility of these data for either research or quality assurance, particularly if patients that undergo the robotic procedure are from a different market segment than the traditional open approach.
With these limitations in mind, my laboratories in Boston and Tucson developed a database of over 300 patients undergoing robotic heart surgery for comparison to an equal number of patients that underwent open chest cardiac surgery. On both the 0-10 pain scale and narcotic requirement, pain was significantly worse on the first day after robotic surgery, with an average pain score of around 7 (vs. 5 out of 10 for the open group) and a narcotic requirement that was 40% higher than open surgery. It seemed counterintuitive, because less invasive cardiac surgery obviously reduces tissue trauma but it clearly did not reduce the experience of pain. At first glance, the Boston surgeon’s criticism was proved to be correct.
Basic science has revealed the way that surgical tissue damage creates the unpleasant sensory experience of pain: tissues innervated by sensory nerves are damaged, local nerve endings are activated and a neural signal is transduced. This signal is transmitted to the brain which then leads to the conscious interpretation of pain. Not all tissue trauma leads to pain. For instance, an endoscopic biopsy of stomach mucosa on an awake patient causes trauma but no pain because this area is not innervated. In contrast, the incisions needed for robotic surgery are made in an area where there are more nerve endings (intercostal ribs and muscles) than if the incision was in the sternum. Pain signals from these areas end up more intense, explaining the paradox of how smaller incisions of robotics could lead to more pain than the large sternal splitting procedure of open cardiac surgery. More pain is limited mainly to the first 2 days. By the 3rd postop day, the inflammation in the intercostal ribs and muscles seems to rapidly resolve and the degree of pain nearly vanishes. Incidentally, the first 2 days is the timeframe that most physicians use to judge the pain caused by a surgical procedure. Thereafter, patients are usually out of the hospital and out of the view of clinicians.
Pain isn’t isolated to a sensory experience. There is a subjective, emotional component that is not adequately assessed by pain scores and narcotic requirements. This is assessed by direct questioning of patients about how they remember their pain control after surgery. Most hospitals obtain surveys of customer service 2-3 weeks after hospital discharge that include questions about pain control during hospitalization (e.g. HCAHPS questions #13 and 14). These survey results are made available for quality improvement purposes, which provided us a convenient way to assess differences between our two surgical procedures. A statistically accurate analysis of the survey results in my two cohorts suggested significantly better pain control after robotic surgery. There was an 85% response that pain was well controlled after robotic surgery (95th percentile) as compared to only 55% (48th percentile) after traditional open surgery.
Compared to alternative metrics, a method of assessing pain by asking patients about their experience leads us to a very different answer about pain after robotic surgery. Use of such surveys to measure pain seems more in compliance with the standard nursing definition of pain as “whatever the experiencing person says it is”. Clearly the pattern of pain after robotics differs from open surgery. Patients are in a better position than clinicians to explain which pattern yields a better patient experience. Accepting the validity of patients’ own memories, it becomes clear that pain after minimally invasive surgery was less troubling than a sternal incision.
These survey results, while encouraging for robotic advocates, still do not establish the final truth on the topic of pain. We must reconcile the differences in pain that was experienced (pain scores, narcotics required) vs. what was remembered (survey results). The difference we noted was dramatic, but equally stark differences between experiences vs. memory of pain has been noted in other studies. In a randomized trial of patients undergoing colonoscopy, half of the patients were allocated to a group that prolonged the procedure by leaving the colonoscope within the rectum without any manipulation for 3 minutes after completing the diagnostic procedure. Even though this group had a mildly uncomfortable prolongation in their procedure compared to the control group, they remembered the procedure as less painful and were more willing to undergo future colonoscopic screenings. They formed this memory because the ending of the procedure was remembered as less painful than one in which the colonoscope was removed as soon as possible. The fact that their procedure was prolonged needlessly in order to produce a period where the pain was less (but still mildly uncomfortable) was routinely ignored.
As these examples illustrate, the task of evaluating postop or procedural pain can be quite ambiguous for patients that aren’t experts and unclear on what aspects to evaluate. For instance, should “credit” be given for periods of sleep or general anesthesia when there is no pain or just evaluate those times when awake? Should an average pain level for the entire hospital course be considered or just the best and worst parts? Without clear instructions, laypeople deal with this type of ambiguity by constructing a narrative and fitting their experience into that narrative. This usually involves a search for a “moment of truth” that depicts the overall pain associated with a procedure. It is this moment – like how they felt towards the end of their hospitalization or procedure – that sticks in their memory. At the same time, other seemingly important aspects are ignored – like all the times they needed narcotics for pain relief or having to endure a colonoscopic exam that was longer than it had to be. A patient’s memory has its limitations. Only the moments that are remembered are available for judging pain control after a surgical procedure.
It may seem counterintuitive that memory works this way. All mammals have evolved a primal fear of pain; misjudging it endangers survival. Why do human mammals routinely create false memories of their pain compared to how things actually occurred? The answer to this contradiction lies in an emotional response to pain that is uniquely human. Pain causes a phase of emotional distress that strongly influences our thinking but usually dissipates quickly over time. Once the emotional fog is lifted, a gap occurs in our ability to empathize or reconnect with what we were feeling and thinking at the time of such an altered state of mind. Memories of postop pain in a more clear-headed state (e.g. towards the end of the hospital stay) are given disproportionate weight, creating a false impression of the overall experience.
My experience with robotic surgery illustrates some notable anecdotes of patient memories that seemed a bit inaccurate. Some of the patients whose initial levels of pain were particularly severe occasionally react with phrases like “I wouldn’t have undergone this procedure had I known about the pain…” or “I had no idea it would be this bad…”. This response is unusual, so I make a habit of noting it in the record and following up with the patient during their return clinic visit 1-2 weeks after discharge. Without exception, these patients lose their emotionally charged views at this follow-up. Many times, the pendulum swings and patients previously screaming bloody murder now say their pain control was excellent after all. They can be reminded of their prior views right after surgery, but their final opinions seem fixed by the narrative that the pain ended well, so as a whole it must have been OK.
Our culture also shows a bias towards how things end. This is how we want live our lives. Long-term happiness comes from forming a favorable narrative of your life. We define a virtuous life as one in which we are at peace on our deathbed, regardless of how we lived the other 25,000+ days of life. A gangster that donates all his loot to charity is remembered more favorably than a man who commits a transgression at the end of a life filled with virtue. It may be good to live life in the moment, but commitment to this advice seems reserved for eastern religions. No surprise that our memory of pain is treated similarly.
Altered memories about pain could also be a side effect of the heart surgery itself. The goal of anesthesia is to prevent memory and awareness. This goal is focused on the intraop period but many of the brain receptors targeted by anesthesia show altered activity weeks after drug is eliminated. Combination with sedative agents given after surgery, such as benzodiazepines, puts patients at risk for antegrade amnesia. Other systemic effects of surgery can provoke memory loss, particularly in elderly patients. Any of these issues can influence the accuracy of memories.
Acute pain causes a myriad of alterations (immune function, wound healing, blood pressure, heart function, glucose metabolism and psychologic function), which obligates physicians to relieve pain, even if it is not remembered or used in any judgments. A multidimensional approach works best, starting preop with setting clear expectations about the amount of pain to be expected. Many patients have the unrealistic idea that minimally invasive is non-invasive and has little or no pain. Pre-emptive anesthesia using thoracic spinal blocks and local rib blocks has been useful at reducing the patient’s sensitivity to pain even beyond the duration of effect of the injected agents. Postoperatively, Precedex infusion is a very helpful way to reduce narcotic requirements in those with severe pain. Those with preserved renal function usually respond well to toradol IV because of its rapid impact on musculoskeletal inflammation and spasm. On occasion, patient controlled analgesic (PCA) devices are helpful to improve comfort and reduce anxiety when the required doses of narcotics is high. Non-pharmacologic approaches may also help. Recently, the concept of mindfulness has been used to focus patients on their experience of pain and all its “fight or flight” sensations. This helps separate the actual discomfort itself from a second layer of suffering caused by its emotional toll. Such an approach often allows subjects to do more in their everyday lives even if pain or psychological distress do not decrease.
Hospitals that are world-class at managing the patient experience (and consultants that guide them) often design their care with the “insider” knowledge that the final impressions are the most powerful. Focusing improvements in those aspects of care that most patients use to evaluate their experiences is undoubtedly an efficient way to improve patient satisfaction. At the extreme, this approach might end up exploiting errors in patient memories as a means of getting better satisfaction scores, which seems a bit sinister. If our goal is the best possible care, there is no substitute for improving actual (not remembered) experiences.
Pain caused by robotic surgery is a perfect example of the type of topic that I like to address in this blog. Everyone desires simple answers to what seem like simple questions. The idea that pain is reduced by less invasive surgery is simple for patients and advocates to understand. At the same time, those skeptical about the value of this procedure note the early increase in pain scores after robotic surgery. The simple answer from skeptics about why there is a discrepancy between pain assessed in the hospital and after discharge is that the latter dataset is unreliable. The truth is that both metrics are needed to provide a complete picture. The future of using robotics in cardiothoracic surgery depends largely on the willingness of us all to pursue more in-depth analyses when needed. Honest discussion of strengths and weaknesses of any novel approach is the key to getting patients and referring providers interested and driving process improvement.