Management of High-risk Patients

The Royal College of Surgeons of England/Department of Health. RCSENG. 2011.

Bottom line: This 2011 document by the Royal College of Surgeons of England and the Department of Health highlights the need for improved perioperative management of high-risk surgical patients who although account for only 15% of surgical procedures, they make up 80% of perioperative deaths. The authors identify numerous practical  recommendations on how to implement improved perioperative care in these patients. 

Major points:

1. Current pathways for emergent/urgent general surgical patients are typically ad-hoc and unstructured. There needs to be a structured pathway for these patients with a focus on early recognition and treatment of emergencies and complications. 

2. These pathways need to ensure that the patient's level of risk need to be appropriately matched with the: (a) urgency in appropriating diagnostic tests, (b) the level of seniority involved in decision making, (c) immanency of surgery and (d) the appropriate level of post-operative care. 

3. Risk calculators such as P-POSSUM should be used (and documented) pre-operatively, and then again post-operatively to provide an accurate estimate of risk. High-risk patients, those with estimated risk of death >5%, should have staff (as opposed to resident/fellow) input on all aspects of their perioperative care. While those at >10% risk should have a staff surgeon and anesthetist present during the entirety of their surgical procedure. 

4. A formalized reassessment of risk should occur at the end of surgery (called an "end of surgery bundle"). All patients >5% should at least receive consideration for critical care, while patients >10% should be admitted to critical care. 

    Whiteman AR, Dhesi JK, Walker D. The high-risk surgical patient: a role for a multi-disciplinary team approach?. Br J Anaesth. 2016;116(3):311-314. doi:10.1093/bja/aev355

    Bottom line: The growing cohort of high-risk, (multi-morbid and often frail and elderly) patients undergoing surgery are the ones that suffer the majority of perioperative complications. Similar to their use most notably in oncology and geriatrics, multi-disciplinary team meetings (MDTMs) should be used to improve perioperative outcomes in these patients.


    Major points:


    1. From a cost-effectiveness perspective, MDTMs should be focused on the small-subset of complex patients in which their is the greatest uncertainty on intervention due to a lack of high quality evidence. In essence, replacing high-quality evidence with the next available option: expert, consensus opinion. 

    2. Pros to MDTMs are: (a) they have been suggested to encourage an evidence-based decision making approach , (b) they streamline care to provide more timely management, and (c) although there is a lack evidence on improvement of clinical outcomes, it is suggested that expert consensus is useful in complex patients whose specific circumstances lack high-quality evidence on intervention. 


    3. Cons to MDTMs include: (a) cost, (b) the role of personalities wherein meeting outcomes favor people who are more forceful with their opinions and (c) high-level of coordination required to arrange a meeting time with specialists who have competing time constraints. 

    4. In the most complex patients, MDTMs should answer the following questions: whether to operate, how to best optimize the patient pre-operatively and which multi-disciplinary team members need to be involved pre and post-operatively. Wherein success would be defined as a reduction in perioperative morbidity and mortality as well as reduction in auxiliary scores such as length of stay.

    Postoperative Monitoring

    The most important aspect of monitoring is recognizing that failure to rescue is predictive of mortality and morbidity. Monitors alone aren’t as valuable as prompt communication and intervention. As a perioperative physician we need to recognize the important of diagnosis and management of perioperative complications

    Ghaferi AA, Birkmeyer JD, Dimick JB. Variation in hospital mortality associated with inpatient surgery. N Engl J Med. 2009;361(14):1368-1375. doi:10.1056/NEJMsa0903048

    Bottom Line: This 2009 NEJM article focussed on the topic of "failure to rescue" - post-operative mortality after a complication. Looking at over 84,000 patients, they found that while the incidence of post-operative complications remained similar across US hospitals, there was a large discrepancy in death following a major complication.


    This article highlights combining the importance of early identification and effective management of complications as an addition to the obvious goal of reducing overall post-operative complications rates as a way to reduce to post-operative mortality rate.


    Major points:


    1. The article looked at a subset of ~84,000 patients (out of a total pool of 363,000. Patients were included if they underwent a general or vascular surgery, and it was an inpatient surgery and if that surgery had a mortality of >1%. While the 83,000 only accounted for 23% of the patients, this high-risk group accounted for nearly 68% of all-deaths among the overall group of 363,000. 

    2. Hospitals were divided based on quintiles according to their post-operative mortality rate. The rate varied from 3.5% in very-low mortality rate hospitals to 6.9% in the highest mortality hospitals. Among these hospital mortality quintiles, there was no statistically significant variation in the rates of post-operative complications. 

    3. The rate of death following at least one major complication - failure to rescue - was statistically significant however. 20.4% in very-high mortality hospitals and 11.4% in very-low mortality hospitals. The biggest differences was seen in stroke patients (46.4% vs 22.5%), deep wound infection (7.1% vs 3.5%) and septic shock (46.2% vs 28.7%).

    4. While most efforts at reducing post-operative mortality are focused on reducing the overall rates of complications. This article highlights that hospitals with similar rates of complications can have nearly a two-fold difference in mortality after a complication. Addressing this concept of failure to rescue is an additional way perioperative physicians can improve perioperative mortality rates.

    5. The authors suggest that key interventions are likely organizational and on the hospital level. They could include improving nurse:patient ratios, optimizing the organization/function of the ICU (such as the utilization of daily rounds led by a certified intensivist) or implementing evidence-based guidelines to target specific complications (such as the Surviving Sepsis Campaign). 

      Posthuma LM, Visscher MJ, Hollmann MW, Preckel B. Monitoring of High- and Intermediate-Risk Surgical Patients. Anesth Analg. 2019;129(4):1185-1190. doi:10.1213/ANE.0000000000004345

      Bottom line: This review article builds on Ghaferi's work and defines "failure to rescue" as "the observation that factors predicting postoperative complications are different than factors predicting overall in-hospital mortality". The article speaks to the benefits and challenges of Rapid Response Teams (RRTs) which were developed as a way to recognize the deteriorating patient early and activate emergency medical treatment in a timely fashion. 

      The authors suggest implementing remote monitoring systems as well as implementing multi-disciplinary team monitoring of post-operative patients (e.g. surgeon + anesthetist) as potential ways to improve the respective arms of the RRT.

      Major points:

      1. RRTs are composed of two arms: the afferent arm is responsible for recognizing the physiological derangements in a deteriorating patient while the efferent arm is then activated and responds with emergency medical management, diagnostic testing and/or escalation of care. 

      2. The key to the afferent arm is "monitoring" or continuously monitoring a patient's well-being in an effort to detect abnormalities in the vital signs which would signify potential deterioration. Challenges include: frequency and accuracy of vital sign measurements, ward staff awareness of a physiologic deterioration, wide variety of Early Warning Systems (EWS), and no EWS that has been validated in surgical patients. 

      3. Regularly training ward staff on the use of EWS has been shown to reduce the time recognizing a deteriorating patient and the time to activation of the RRT. Curiously, studies show that respiratory rate was the single vital sign most closely associated with clinical deterioration but it was also the one the nursing staff put the least importance on. Blood pressure meawhile is one that is given the most importance but is often inaccurate and usually a late sign of clinical deterioration.


      4. Similarly, the use of remote monitoring has been shown to also reduce the time to recognizing deterioration and activating the RRT.  The only caution being that frequent false alarms in an automated system leads to alarm fatigue and can cause a malfunction of the entire response system. 

      5. The efferent arm is more difficult to study and most evidence is based on low-moderate quality evidence. The MERIT study showed no overall effect on unplanned ICU admissions, cardiac arrest or death in hospitals who used RRT compared to those who didn't. However, it was a poorly powered study so it should point to a lack of evidence as opposed to a lack of efficacy. Since RRTs are now standard of care in hospitals, it would be unethical to do a randomized controlled trial. 

      6. Anesthesiologists are uniquely positioned to be part of a multi-disciplinary team (with the surgeon) to provide early recognition of deteriorating patients and recommending further management. Anesthesiologists are trained to identify early physiological derangements and are also used to working with surgical patients. The TRACE study is hoping to show that routine visits from anesthesiologists to post-surgical patients will improve 30-day mortality. 

      Jones DA, DeVita MA, Bellomo R. Rapid-response teams. N Engl J Med. 2011;365(2):139-146. doi:10.1056/NEJMra0910926

      Bottom line:​ With emerging evidence of "failure to rescue", Rapid Response Systems (RRS) have become widespread in the US, Canada, Australia and Europe. They are based on the principle that early interventions improve outcomes in almost all emergency medical situations. RRS strive to bring critical care to the patient's beside (an ICU without walls) when critical deterioration occurs although there is presently a lack of Grade 1 evidence for the efficacy of RRS. This 2011 NEJM article informs the reader about definition and composition of a RRS, their history and evolution, the evidence supporting their use, the controversies surrounding them, as well as the potential benefits and limitations of such systems. 

      Major points:

      1. The effectiveness of RRS are based on 5 single-centre before and after studies which showed a reduction of rates and a "dose-response" curve (the greater the assessments per 1000 admissions the greater the benefit). However, the MERIT study, a major multi-center cluster-randomized controlled trial, failed to show efficacy. It was however limited by its statistical power among other controversies so it is difficult to draw conclusions from. Subsequent meta-analyses have also raised questions as to whether there is a benefit. 

      2. Most common conditions associated with failure to rescue are: acute respiratory failure, acute cardiac failure, changes in consciousness, hypotension, arrythmias, pulmonary edema and sepsis. Regardless of the condition, early warning signs for clinical deterioration are seen in the vital signs. Vital signs on the ward however, are typically infrequently, inaccurately or incompletely recorded. Furthermore, it is unknown what proportion of hospitalized patients have abnormal vital signs but do not go on to experience a serious adverse event. 

      3. Implementation of RRS can take up to a year and require support from hospital leadership, senior medical personnel as well as senior nursing personnel. Emphasis should be on providing a quick second opinion as opposed to taking over care. The target should be >25 calls per 1000 admissions and ideally >40 at mature academic centres.

      4. Controversies surrounding RRS include: lack of evidence, their cost, the optimal composition of the team remains unknown, potential to "de-skill" ward staff, diversion of care and hence jeopardizing ICU patient safety. 

      Story DA, Shelton A, Jones D, Heland M, Belomo R; Austin Health Post-Operative Surveillance Team Investigators (POST). Audit of co-management and critical care outreach for high risk postoperative patients (The POST audit). Anaesth Intensive Care. 2013;41(6):793-798. doi:10.1177/0310057X1304100616

      Bottom line: This single-center Australian study implemented a post-operative surveillance team (POST) comprised of a physician and ICU nurse to round on high-risk surgical patients on two surgical wards for five days post-op or until discharge. Compared to controls, The POST group had no significant differences in mortality (~2.5% for both groups) while having a 2 day longer length-of-stay. 


      Major points:


      1. The POST group comprised of 185 patients over 4 months. Patients were included if any of the following was true: (a)  aged 55 or older and not electively admitted, (b) aged 80 or older and electively admitted (c) admitted to ICU before going to ward or (d) identified by the ward team for anything other reason. 

      2. In both the control and POST group, the hospital's two main rapid response systems still remained operational: Medical Emergency Team (MET) and ICU nurse outreach. The MET team was called upon when ward patients met certain changes in physiological criteria. The ICU nurse outreach team rounded on patients discharged from the ICU and on patients after a MET call. 

      3. There were 194 POST patients and 1185 in the control group. Control patients were on average 2 years older and had more emergency surgeries while POST patients were more likely to have pancreatic, colorectal, major HBP or upper GI surgery. POST patients remained in hospital 2 days longer. No difference in MET calls during the hospital stay or during the first five days post-op. There was no difference in mortality (2.5% for both groups). There was no difference in the time to first MET call post-operatively. 

      4. The authors postulated that their was no benefit to POST for any of the following reasons:

      (a) POST is not an effective intervention 

      (b) POST is effective but was not implemented properly. More senior staff should have been used or patients should have been monitored for longer than 5 days.

      (c) POST was unable to add value on top of the existing RRS modalities already in place at the hospital (MET and ICU nurse outreach).

      (d) LOS might not have been an appropriate outcome to study as greater surveillance could logically prolong LOS as problems are actively managed. 

      Ahmad T, Bouwman RA, Grigoras I, et al. Use of failure-to-rescue to identify international variation in postoperative care in low-, middle- and high-income countries: a 7-day cohort study of elective surgery. Br J Anaesth. 2017;119(2):258-266. doi:10.1093/bja/aex185

      Bottom line: This 7-day cohort study looked at the outcomes of 44,814 patients across 27 low, middle and high-income countries. Despite an overall failure-to-rescue rate of 2.8%, further analysis showed a two-fold difference in the best and worst performing hospitals (3.6% vs 1.7%) suggesting the presence of preventable postoperative deaths. This study supports the use of failure-to-rescue (death after post-op complication) as a metric to compare the performance of healthcare systems at the international level and also stresses the importance of high-quality perioperative care in reducing preventable deaths.

      Major points:

      1. The overall mortality was 0.5%. This varied from 0.6% in the quintile of hospitals with the lowest surgical volume (Q1) to 0.2% in the highest surgical volume hospital quintile (Q5). 


      2. Overall, 17% of all patients developed a post-operative complication. 12% of all patients developed only a single complication, while 5% developed two or more. The median hospital stay was 4 days in patients with no complications and 8 days in those who developed a complication. 

      3. Interestingly, hospitals with the highest complication rates did not have the highest failure-to-rescue rates. This suggests a difference in abilities of hospitals to identify complications and escalate care of deteriorating patients in a timely manner. 

      4. As global initiatives strive to increase surgical access to all parts of the world, it is crucial to take into account the importance of effective perioperative care in the prevention of post-operative surgical deaths. 

      Michard F, Sessler DI. Ward monitoring 3.0. Br J Anaesth. 2018;121(5):999-1001. doi:10.1016/j.bja.2018.07.032

      Bottom line: The author compares & contrasts the different levels of ward monitoring and the strengths and limitations of each highlighting wearable, wireless monitoring (ward monitoring 3.0) as a future direction to decrease post-operative morbidity and mortality. 

      Major points:


      1. "Ward monitoring 1.0" is the traditional intermittent 4-6h nursing vital sign checks. Studies show that ward hypoxemia lasting 1hr goes unnoticed in 1/3rd of patients, and in 10% of patients lasting 6hrs. Nurses checking vital signs every 4-6 hrs missed 90% of hypoxemic episodes that lasted 1 hrs. Similar findings for hypotension which is worrisome as post-operative hypotension is associated with a three-fold increase in MI risk. 

      2. Ward monitoring 2.0 is traditional, continuous bedside monitoring with tethered/wired monitors for vital signs. Studies have suggested that continuous bedside monitoring improves outcomes compared to intermittent monitoring. The author highlights studies which show fewer rescue events, critical care transfers, number of cardiac arrest calls and decreases in hospital mortality when continuous beside monitoring is implemented. 

      3. Ward monitoring 3.0 is described as medical-grade sensors that are wireless and wearable and make physiologic monitoring practical for ward patients. They can be combined with emerging technologies such as machine learning. More research is needed to determine


      (a) Whether they can accurately monitor vital signs (i.e. distinguish real derangements from artefact)? 

      (b) Whether excessive false alarms and resultant alarm fatigue can be avoided? 

      (c) Can they be used to detect clinical deterioration early enough so that an intervention can be applied?

      (d) Can they ultimately decrease the rates of serious adverse events?