Every day, millions of patients undergo procedures in ambulatory surgery centers (ASCs), and a very small percentage of them have complications during and after the procedure. Fortunately, new technologies promise to address problems that have challenged patients and their surgical teams.New analytical tools using artificial intelligence and machine learning, coupled with data streams from patients, can be used to better understand patient risksand develop treatment strategies to reduce complications and improve outcomes.These new tools promise to reduce waste (defined as care that does not add to the patient’s health).For example, a patient who develops a wound infection post-operatively that requires re-operation and hospital admission may benefit from those interventions, but if the patient had gotten treatment earlier for the wound infection,hospitalization and re-operation could have been avoided altogether. Waste is common in our healthcare system today, but things are already starting to change.
How can new technologies reduce waste and improve outcomes for patients treated in ASCs? These tools can help predict risk of complications, better manage surgical procedures, and post-operative recovery.Physicians and teams managing surgical patients face several difficult issues: when is it safe to take my patient to surgery? What are the identifiable risks of surgery for this patient and what can be done to minimize those risks? How can the surgery be made as safe as possible? What does this patient need to heal successfully after surgery and what can be done to minimize surgical complications that require treatment? Better answers to these questions lead to better outcomes for patients and reduce wasteful expenditures.
Consider the example of a 47-year old woman with a history of high blood pressure managed by diet and exercise, who presents to her primary care physician complaining of several days of pain in the upper right part of her abdomen, along with back pain that radiates to her shoulder blade.The pain gets worse after she eats. Physical examination and imaging studies are consistent with mild inflammation of her gall bladder, but it does not appear to be acutely infected.The patient is referred to a general surgeon for consultation, and the surgeon recommends removal of the gall bladder in a few weeks.The patient is scheduled to have the procedure done through a laparoscope, since the recovery is much less difficult and painful than an open operation.
How are surgical risks assessed for this patient?Traditionally, the patient goes to a pre-operative clinic where she is evaluated for surgical risks, based on a history, physical, and any labs or imaging studies needed.If necessary, medications or treatments are prescribed to reduce the identified risks, and the patient is asked to consent to a procedure knowing that there is risk.Using new technologies that rely on evaluating broader, larger data sets can help risk-stratify the patient with more certainty. In addition to understanding the patient’s clinical exam and lab results, medical and social history, artificial intelligence systems can take data from a wide variety of sources, including medical records, claims systems, census data, credit history, genomic data and other information to identify specific risks for this patient. These tools can be used to risk stratify the patient and suggest any needed treatments to reduce risk and design a post-operative care plan that may include a visiting nurse or other home health services. These tools can also use machine learning to evaluate the impact of its recommendations and make better, more accurate risk stratification decisions in the future.
In the case of the patient undergoing gallbladder surgery, the pre-op analytical tool can recommend home health services and suggest ordering medical nutritionbecause the patient lives alone and has limited help at home. In the future, it is likely that this patient would receive a patch or wearable sensor array that could track her vital signs and identify changes or variability signaling potential risks. For example, if she began to run a low-grade fever, that could be identified and brought to the attention of her surgical team well before she checked in for her surgery, which would allow for earlier intervention and treatment, and potentially avoid surgical complications.
Use of analytics tools during surgery can also reduce risk.The development of software as a medical device (SAAMD) has enabled companies to create algorithms for taking in multiple data feeds and generate warnings, alerts, and prioritize among the alerts so the anesthesiology staff can address problems earlier and see built-in guidance for how to manage the risk identified and reduce the likelihood of bad outcomes.Use of tools like this will continue to expand as new data feeds become available, and real-time, minute-by minute monitoring becomes more common, even outside of operating rooms and step-down units.
Another technology, surgical robots may have an impact because these systems generate multiple streams of surgical data, an important aspect that adds to the value of these systems.Unlike traditional surgery, robotic surgery generates multiple data feeds, ranging from operating time of the surgeon, to what surgical approach was used, video, and information on the “dexterity” of the surgeons.The data is valuable for real-time analysis of operating surgeons, which can be done through remote views of the robot’s cameras and other data, and also for compiling statistics on surgical performance. For a low-risk 47-year old woman undergoing gall bladder removal, understanding which surgeons have had the best results is important data that is not currently available. The day may soon come when “smart robots” will be able to recommend surgical approaches and sense early problems, like uncontrolled bleeding.
The post-operative period also has specific risks that can be identified earlier and better managed through analytics to predict needs, and wearable sensors to identify changes in clinical conditions that can be managed earlier before they require emergency treatment. For the gallbladder patient mentioned above, a smart pre-op system would have identified potential gaps in her post-discharge needs, such as help with transportation, and access to medically-healthy, low-fat meals tailored to meet the needs of a patient after a gallbladder surgery. Placing a disposable sensor patch to measure vital signs and physical activity would also help to identify issues. It will be possible to use the same software that was used pre-op, that had learned her usual vital signs and activity patterns, to sense changes or variability that could signal onset of post-operative complications. The patient could also use a voice-activated smart agent, like Amazon’s Alexa or Apple’s Seri to get help or ask questions about symptoms. In fact, “intelligent agents” like Alexa and Seri can be programmed to ask questions at pre-set intervals, and trigger outreach and follow up by clinicians if needed. This is helpful for issues like pain management, which are not easily quantified or captured using sensors. These tools are also helpful for combining data on medication adherence, vital signs, and input from intelligent agents could provide regular, real-time reporting to clinicians and identify opportunities to intervene early and avoid emergency room visits, hospitalizations, as well as other wasteful, potentially-preventable expenditures.
The ASC of the future will be enabled by analytic technologies, sensors, and software that interprets data flows and learns patient-specific patterns and deviations that signal changes in clinical condition.These technologies will lead to better, safer careand reduce wasteful expenditures on hospital care and emergency room visits.