Former SCOAP Medical Director Dr. David R. Flum wrote a few columns that appeared in the monthly newsletter to SCOAP supporters, but you can view them all in one place right here on our website!
In these columns, Dr. Flum answered questions that have come up from members of our community by referencing articles or abstracts that help to provide more information:
Do you or your colleagues have a question YOU would like to see answered here and in the SCOAP newsletter? Just send us an e-mail…
Why are we focusing on 12 negative lymph nodes in patients with colon cancer given the recent evidence?
Some surgeons have balked at this metric because getting at least twelve negative LNs evaluated is a group activity. Surgeons remove the specimen but the pathologist has to know to look, and then find at least that many lymph nodes. Most surgeons believe they have been removing the same, appropriate and complete amount of mesentery for years and that if there are 12 lymph nodes in there then they should be found. While there are special specimen preparation techniques for identifying more lymph nodes, at most SCOAP hospitals the rates of 12+ negative lymph nodes have been increased simply by letting the pathologist know that this is important for post-surgical treatment and that their performance of this metric is a target of a QI effort. That constant feedback of information to pathologists seems to have driven up the metric considerably. At some SCOAP hospitals they have moved to a “templated pathology report” that has the pathologist having to “opt out” of 12+ negative LN if they COULD NOT identify them. These templates remind pathologists that getting 12+ negative LNs is important. This is one of the reasons SCOAP hospitals are now some of the best performers on this metric. The 12+ negative LN metric was being considered by the National Quality Forum as a quality metric that would be involved in pay for performance programs. The enthusiasm for the 12+ negative LN metric for that purpose faded a bit with an article in (Wong et al JAMA 2007) that found that survival was similar at hospitals with high and low performance on the 12+ negative LN metric. This was a complicated article but there are a couple of salient points that explain the motivation for SCOAP using this as one of its targets. In that study of over 30,000 patients in the SEER registry;
1. Patients who had at least 12 negative LNs removed/examined had a much lower risk of death over the study timeline (17% lower overall and 31% lower for Stage 2 cancers where you would expect correct staging would matter and 11% lower among Stage 3 cancers where appropriate treatment with chemotherapy would matter). This finding-12+ negative LN means better stage for stage survival and overall survival- has been consistently identified.
2. The use of chemotherapy did not vary between hospitals who were high and low performers of the 12+ negative LN metric. In other words, there was underuse of chemotherapy in all hospitals, regardless of 12+ negative LN status and so the expected benefit (more chemotherapy) of more correct LN staging was not being recognized.
3. There was general underuse of the 12+ LN metric among the hospitals examined with even the “best” hospitals using it inconsistently. In the study the highest quartile of hospital performance on this metric was 61%. Compare that to how well we’re doing in Washington State-on average SCOAP hospitals performed this metric 71% of the time (in the last year) with several hospitals getting it right all the time.
In short the authors found that the 12+ negative LN metric was an ineffective way to judge the quality of a hospital but still an important goal for any given patient. Remember though, SCOAP is not about ranking hospitals high to low, it’s not a report card. Rather, we’re trying to create a tide that lifts all boats. Every patient with colon cancer should have the right number of LNs evaluated so that a correct determination of their metastatic status can be determined and that they can get the right adjuvant care if needed. By spending some time focusing on this metric, getting surgeons and pathologists talking and helping hospitals use best practices in the pathology department, SCOAP is helping all patients with colon cancer have the best chance at survival.
Why are we focusing on glucose control in general surgery given the recent evidence?
The evidence for blood sugar control to normal levels is emerging. The strongest evidence for perioperative blood sugar control comes from the cardiac surgical literature where it has been found to dramatically reduce the risk of infection in multiple randomized trials. Although the randomized trials in general surgery have yet to be performed a recent cohort study (Ramos et al. Ann Surgery 2008) of nearly 1000 patients undergoing vascular and general surgery found a 30% increases in infection for every increase of blood sugar by 40 Units above 110 mg/dL. Some surgeons have reported concern about some news they heard regarding sugar control in patients after surgery. They may have been thinking about a recent study by Arabi et al (Critical Care Medicine 2008) that randomized patients to VERY tight glucose control (80-100) or strategy that kept sugars controlled to less than 200. The very tight sugar control did not improve outcomes and caused many more episodes of low blood sugar. The authors concluded that this level of tight glucose control was not beneficial.
It’s very important to frame these data as we try to make Washington State surgery as safe as it can be. No one thinks that blood sugars above 200 are good for patients, and most think they increase infections. So how do we make it so hyperglycemia does not hurt our patients? The SCOAP metrics are really keyed into checking the blood sugar and then starting a response with insulin if it is high. How tight doctors at your hospital should control glucose is a matter of debate, but there is general agreement that you should measure AND control extreme values. As a performance metric, SCOAP reports back to you how often insulin is started in patients with blood sugar over 200. That should help raise awareness and target underperformance.
So why does the SCOAP Surgical Checklist include a prompt to start insulin for blood sugars greater than 125? The SCOAP Surgical Checklist is intended to catch errors before they occur, so the lower blood sugar prompt was selected because of findings in studies showing that blood sugar greater than 125 increases the risk of infection. It’s not that infection risk starts at 200, just that 200 is an extreme value. Our goal with glycemic control in general/vascular surgery is to reduce infection while making sure we don’t hurt patients with glycemic control that is too tight. This suite of insulin orders related to all phases of perioperative care and handoffs may help hospitals that are working to address perioperative hyperglycemia in the safest way.
For now, remember:
- Use the SCOAP surgical checklist as a prompt to get blood glucose on all patients with diabetes who are having an operation-this is a SCOAP metric.
- For patients with elevated blood sugars, have the anesthesia team start insulin and check the blood sugar after starting to titrate to an appropriate level (best if done using an algorithm). Proportion of patients with blood sugar over 200 who are started on insulin is a SCOAP metric.
- Using the SCOAP surgical checklist to encourage the anesthesia team to start insulin when it’s elevated will mean controlling blood sugars before they ever get into the highest range. Starting insulin at 125 is not a SCOAP metric, but it’s a good idea if done safely.
For further reading: Tight Glucose Control May Increase Mortality in Critically Ill Patients (via MedScape)
Why are we focusing on Beta Blockers given the current evidence?
The SCOAP metrics regarding beta blockers are emerging with the changing evidence. The POISE trial (POISE Study Group. Effects of extended-release metoprolol succinate in patients undergoing non-cardiac surgery (POISE trial): a randomized controlled trial. Lancet 2008. DOI: 10.1016/S0140-6736(08) 60601-7.) demonstrated an increase in complications associated with beta blocker use in non-cardiac surgery after several other RCTs had suggested a benefit. A recent meta-analysis found little benefit to starting beta blockers among at risk surgical patients. While the POISE trial has been widely reinterpreted to be the results of a much higher than normal dose of beta blockers, these studies have inserted a note of caution to a growing interest in getting more patients with cardiac risk factors started on beta blockers before surgery.
SCOAP has always focused on avoiding acute beta blocker withdrawal in patients already taking beta blockers. This is currently the ONLY metric in SCOAP reporting related to beta blockers. The SCOAP Surgical Checklist asks if there is a plan to continue beta blockers after surgery as a way to avoid the 4.5 fold increased risk of a heart attack that comes with beta blocker withdrawal (Psaty et al JAMA 1990). Our colleagues at Harborview asked whether the real metric should be avoidance of an elevated heart rate in someone on beta blockers. While we agree that avoidance of tachycardia is the real goal of beta blocker continuation, tracking on heart rate is challenging, non-standardized, and does not meet the emerging metric that is coming through CMS’s SCIP initiative.
So why is SCOAP also collecting information on cardiac risk and use of beta blockers before surgery? For one, we have to know who is taking beta blockers before surgery so we can know who has them discontinued afterwards. Secondly, SCOAP hopes to be part of developing the emerging evidence on beta blocker use in general surgery so we gather data on patient cardiac risk (this is also helping in developing risk adjustment for outcomes) and cardiac outcomes after surgery. While in an earlier version of SCOAP reporting we had been listing “use of beta blockers among high risk patients” you will no longer see this metric as the evidence is under development. It is likely that someday the variables that constitute sufficient levels of high risk for heart attack will emerge and that beta blocker use may be a part of the MI prevention strategy in patients who are high risk. SCOAP should help inform this discussion as the evidence matures.
For additional discussion regarding this issue, please review the following article:
Lisa Nainggolan. New meta-analysis of beta blockers in noncardiac surgery: Time to abandon their use, or not?. theheart.org. [Clinical Conditions > Interventional/Surgery > Interventional/Surgery]; Nov 12, 2008.
For now, remember:
- Restart beta blockers on patients who were taking them before surgery.
- Document if there are contraindications to continuing beta blockers such as slow heart rate, hypotension, or if the patient is taking beta blockers for a non-cardiac reason (e.g., anxiety). This will help make sure that your SCOAP metrics are more informative.
Why are we focusing on use of CT and US in patients undergoing appendectomy?
SCOAP includes 2 metrics related to appendicitis. The first relates to the accuracy of CT and US in patients with presumed appendicitis who are taken to the operating room. Accuracy is not a very accurate word however – what we are really measuring is concordance of operative findings and radiologic findings. This is neither sensitivity nor positive predictive value but it is the best we can do given that SCOAP does not track patients who had negative or positive scans who DID NOT undergo an operation. SCOAP is also tracking how often women of reproductive age get some form of advanced imaging. Earlier work revealed that when women in this age group DO NOT get imaging, the negative appendectomy rate is as high as 45%. Statewide estimates before SCOAP (using administrative data) showed the negative appendectomy rate for this group was 25%. These very high rates of unnecessary appendectomy can be reduced to levels less than 10% with the advent of CT and US.
While there is an emerging debate about the amount of added risk a spiral CT scan confers on the development of cancers (Einstein et al. JAMA 2007) there is little consensus about the absolute risk of diagnostic imaging and no risk to ultrasound. SCOAP’s focus is not on getting more CT scans but on getting more accurate scans. We encourage SCOAP hospitals to engage the radiologists involved in the interpretation of these scans in a QI activity surrounding negative appendectomy and diagnostic accuracy.
Lastly, we’ve had questions about whether the surgeon or pathologist’s interpretation of the condition of the appendix is being used to determine appendicitis or perforation. A study of surgeon’s macroscopic view of appendicitis found that when they think it is appendicitis from the look of the appendix they are usually right, but when they say it’s NOT appendicitis they are “right” (as determined by the pathologist) less than 50% of the time. Similarly, perforation appears to be more reliably determined by the pathologist. We use the pathology report to determine both metrics.
For now, remember:
- For women of reproductive age, the rates of negative appendectomy are VERY HIGH when no US or CT are used. Our experience with SCOAP data encourages the use of these tests among women of reproductive age. The recommended use of these tests must be linked to…
- The accuracy of CT and US should be acceptable if they are to be used in diagnosing appendicitis. SCOAP benchmarks of greater than 95% accuracy are being achieved by many hospitals and should be a target for all hospitals.
- An inaccurate diagnostic test does no one any good, can delay operation, and is a waste of money. The test accuracy can be VERY high when proper technique and interpretation is applied.
Why does it seem that SCOAP wants everyone to get a CT scan given recent evidence about radiation’s effect over time?
SCOAP does not want everyone to get a CT. SCOAP is made up of surgeons like you who looked at data on negative appendectomy and, not surprisingly, noticed that we were doing really poorly at diagnosing appendicitis among women of reproductive age and young children and have been working to do something about it. The rate of NA among women was nearly 1 in 4 prior to SCOAP. The answer is not necessarily getting more imaging but among this group that’s highest at risk for NA the answer may be using ACCURATE diagnostic imaging (US or CT). While early reports about CT accuracy led us to believe that the scans were nearly always accurate, we all know that the scan quality can vary widely – and an inaccurate scan is no help. In SCOAP, focusing on increasing scan accuracy through benchmarking has changed the quality of imaging in the state. Use of accurate scans has brought down the rate of NA among women to less than 10%, and the average SCOAP-wide rate for all patients is now approximately 7%. At hospitals that have been most successful, surgeons, ER docs, and radiologists have worked together to increase accuracy of CT and US and have then used these tests for ALL women of reproductive age. That seems to make the most sense. The data are still out on the long-term impact of a single CT scan on women of reproductive age, but we should probably emphasize the use of accurate US as our colleagues at Seattle Children’s Hospital do (90%+ use of US compared to the rest of the SCOAP hospitals that are 90%+ CT). There seems to be little benefit of routine CT in men although even among men in SCOAP, those getting imaging have a lower risk of NA compared to those without imaging. The rate of NA among men is so low that the relative benefit might not be worth it. At most hospitals, it will be most beneficial to benchmark and improve the accuracy of the CT scans. This can be done by feeding information back to radiologists, meeting with them to understand their process for QI, suggesting best practices from the literature (SCOAP can help with those), and because so many hospitals have asked for this, we are including information about contrast use and route into the dataset.
We have two examples of how imaging accuracy data has changed practice. At one of our teaching institutions, the July-September quarter of data showed a big dip in accuracy compared to the prior quarter. This is probably the result of new fellows interpreting the scans. The SCOAP data are being used as a guide and monitor to prevent lapses in quality as the department prepares for next year’s radiology trainees. At another, smaller hospital, lower than benchmark accuracy data were used in a negotiation with the outsourced radiology vendor, and this benchmark incorporated into the metrics for further contracting.
SCOAP hospitals have some of the lowest rates of NA in the nation, but this has not come at the expense of a higher perforation rate. The perforation rate has not increased while NA has come down. Many now consider perforation to be a pre-hospital event caused by delays in accessing the healthcare system. Because several surgeons have asked for more information about perforation at their institution, we’ve added some new metrics about timing so that the issue of in-hospital delay can be better assessed and benchmarks established. This should also shed some light on the long-standing debate about whether patients with appendicitis are hurt by waiting until the morning for an operation. Many surgeons have asked whether this practice of delay is endorsed by the SCOAP data and we have not had the information to answer the question. SCOAP should have lots to add on the subject as we’re the only state in the nation with all appendectomy cases under surveillance.
SCOAP’s progress with appendicitis was just reported in the Annals of Surgery with a recent letter to the editor (2008 Oct;248(4):557-63).
Lastly, some hospitals that have been working on appendicitis since SCOAP started are now starting Vascular and Higher-risk (esophagus, pancreas and liver) SCOAP modules and have asked if they can focus less on appendectomy. SCOAP is a voluntary initiative, but it means nothing if we don’t have reliable benchmarks because of picking and choosing which procedures a hospital evaluates. What makes sense is the idea of SCOAP as a roving spotlight that focuses with a sharper beam on some procedures in any given quarter than on others. For those hospitals that have done appendectomy for at least one quarter, that have a lower than average NA rate, and that are taking on another module (such as higher–risk, lower-risk or vascular SCOAP), a sampling strategy keeping a “low-intensity” focus on appendectomy is available. This sampling strategy would evaluate 1-in-5 appendectomy cases to monitor for consistent performance during the more intense data collection of the new module.
Why is SCOAP focusing on blood transfusion after surgery?
Some know that it was surgeon George Crile back in 1906 who administered the first transfusion – and it seems we’ve been arguing about their value ever since. The surgeons who developed SCOAP wanted to track on transfusions so that they could better understand the benefits, risks, and use patterns of this limited resource. SCOAP data reveal widespread variability in the use of transfusions after elective colon surgery across hospitals and wide variation in the amount of blood transfused when patients do get blood. Blood products are limited, cost money, have their own associated risks and for most patients may not add much in a positive way to outcomes. To answer the SCOAP communities’ questions about the “ins and outs” of transfusion we went to the literature for the most recent evidence.
We all agree that fixing critical anemia after surgery is important. Severe anemia can result in inadequate O2 delivery, can increase mortality risk1, and we do know that RBC transfusion improves O2 delivery. The trick is to understand what critical anemia is for your patient. A transfusion trigger is the point at which the risks of decreased O2 carrying capacity exceed the risks of transfusion. There has been considerable debate in post-surgery and critical care environments about the role of a restrictive strategy for transfusions compared to a liberal transfusion strategy. There have been over 12 RCTs on the topic and, in general, restrictive approaches (transfusion threshold between Hb 7-8 g/dL) reduce the number of RBCs transfused, and a restrictive strategy is not associated with adverse clinical outcomes2. Perhaps the best evidence for this comes from the ICU literature, and the TRICC Study3 may be the best of these studies. TRIC was a multicenter, prospective, randomized study of 838 patients in 25 centers with > 24h ICU stay expected and a Hb < 9.0 g/dL within 72 h. The investigators followed one of 2 strategies: 1) Restrictive – maintain 7-9 g/dL or 2) Liberal – maintain 10-12 g/dL. Researchers found a 52% reduction in transfusions and a lower risk of in-hospital death in the restrictive group. In fact, considering higher risk groups (e.g., low APACHE scores and younger patients) the number needed to treat to save one life by using a restrictive strategy was 12. This lack of benefit with liberal use of transfusions suggests we offer little benefit with them. This has not been as rigorously studied after surgery but is the focus of an ongoing investigation in at-risk patients having hip surgery in the FOCUS trial4.
Given the increased risk of death seen in the TRICC study, and given emerging evidence on their risk, some have asked if transfusions aren’t really deleterious for our patients. One of the potential downsides of a transfusion is Transfusion-Related Immuno-Modulation (TRIM). TRIM5 includes multiple immunologic effects of an allogenic blood transfusion including:
- Decreased T-cell proliferation;
- Decreased CD3, CD4, CD8 T-cells;
- Increased soluble cytokine receptor (sTNF-R, sIL-2R);
- Increased TNF-alfa;
- Increased suppressor T-cell activity;
- Reduced natural killer cell activity;
TRIM can result in decreased host defense and immune surveillance, is pro-inflammatory, and numerous observational studies and a meta-analysis have suggested a relationship between peri-operative allogeneic blood transfusion and postoperative bacterial infection6. While some have claimed that these effects are because of “older” transfusions, an RCT looking at this found no benefit to more recently donated units7.
Perhaps of greatest importance to our patients however is that transfusions around the time of cancer surgery are associated with a 42% increase in the risk of cancer recurrence8 – likely linked to some of these immunologic mechanisms, though not necessarily improved through allogenic transfusion. This is particularly relevant in SCOAP where many of our patients undergoing elective colon resection carry the diagnosis of colon cancer. Sometimes when we have done a great operation for cancer and our patients return with a recurrence a year later we throw up our hands and wonder what we could have done better. For many of us who don’t think of transfusions this way, it is possible that one of the factors influencing recurrence rates is how restrictive we are about transfusions.
Based on the available evidence, SCOAP surgeons and hospitals should consider the following:
- If hemoglobin < 6-7 g/dL, transfusion is often necessary;
- If hemoglobin is between 7 and 10 g/dL, transfusion may not be necessary;
- If hemoglobin > 10 g/dL, transfusion not indicated.
SCOAP hospitals that have been effective in reducing the use of transfusions have done so by instituting transfusion triggers (such as the above) as a policy, having surgeons default to “no transfusions for Hg<7″ unless they indicate that it is clinically necessary and work to educate surgeons and staff about the risk and benefit of transfusion. To better inform SCOAP surgeons, this quarter we are expanding SCOAP’s metrics to include the lowest hemoglobin prior to transfusion. This should provide more information about how surgeons at different hospitals use blood products and whether transfusion triggers are really working. While the evidence is still out on what is the “right number” of transfusions after general surgery, it seems that surgeons at some SCOAP hospitals where this is being worked on are doing a better job than others at controlling the use of transfusions. Over the next few years, SCOAP is hoping to provide surgeons with more information about when transfusions are helpful and when they just add risk and cost to our patients. In the meantime, here is what you can do:
- Share with your colleagues the evidence of increased infection and cancer recurrence with transfusion and the lack of evidence of benefit of transfusions.
- Find out if your hospital has a transfusion committee that is working on this issue.
- Have this committee do a quarterly review of postsurgical transfusions so you can understand the reasons why your hospital uses more transfusions than others.
- Volunteer to be part of a group that uses a transfusion trigger with “clinical condition exception rules.”
- Through SCOAP, track improvements in the use of transfusions in elective surgery.
Why is SCOAP looking at ENTEREG?
Entereg is a relatively new drug that blocks the gut receptors for opioids and is being marketed to prevent ileus caused by postoperative narcotics. The drug has been the subject of several randomized trials, with all but one suggesting a reduction in length of stay (LOS). A recent meta-analysis(Archives of Surgery, Nov 2008) suggested a 20 hour reduction. Since we cannot predict who will have an ileus after colon surgery, and since the drug needs to be given before the opioid and through to discharge, those who believe in the drug may find themselves giving it to everyone having colon surgery. Like all things, the drug has a price – but so does an increased LOS because of an ileus. Depending on where you practice and the deal your pharmacy signed with the company, the cost may be as high as $800, and that money counts against the DRG for your hospital’s profitability (increased LOS may also get in the way of profitability but not as directly). Because the differences in LOS coming out of the RCT were less than a day, skeptics have wondered what the effect will be in the “real world” where people stay in the hospital after surgery for many reasons that have little to do with ileus.
ENTEREG is a great example of the power of SCOAP to figure out what we should be doing for our patients based on the successes of our colleagues. If SCOAP is about figuring out best practices and getting all to learn from each other, then ENTEREG may be the test case. Several hospitals are now using ENTEREG, and as soon as the drug came on the market we incorporated it into SCOAP (which is also collecting data on LOS). Measuring ileus is problematic and to do it well is very expensive. Because we wanted to learn as much as we could about the drug to help you make decisions about prescribing it, we added day and hour of first order for solid food – a non-standard measure of ileus but one that can at least be standardized within SCOAP. We should know soon if LOS decreases with ENTEREG use or, if not, whether time to feeding is at least shorter (so that non-ileus related LOS issues can be tackled by SCOAP surgeons). We also met with the company that makes the drug to help support data collection at your hospital about finer measures of ileus measurement, but they were not interested in supporting SCOAP. Expect soon our first data on ENTEREG in SCOAP. If it works, maybe we all need to think about it. If it does not, maybe that is money better spent on other things. In any case, SCOAP is being responsive to your interests and needs as surgeons and hospitals trying to watch the bottom line while delivering highest quality care.
If there are other expensive drugs like ENTEREG (like long-acting Zofran [Aloxi], the subject of an upcoming column) or other interventions you think we should add, please contact us at firstname.lastname@example.org).
What is SCOAP doing with data on length of stay, and are we looking at why patients stay in the hospital longer? Does SCOAP really think this is a metric of quality?
As surgeons, we know there are many reasons our patients stay in the hospital, and we each have our own sense of when a patient is cleared for discharge. This varies considerably based on training and prior experience as well as confidence that the patient can get back to the hospital should there be a problem. Some of us insist that there be evidence of full gastrointestinal function or that a high risk window for a complication (such as an anastomotic leak) be passed. Patients may have complications or suspicion for complications that require additional care, family and social support may not be ready to receive the patient at home, and/or they may have difficulty leaving the hospital because of transportation issues (e.g., distance from hospital). For some patients, insurance status is a co-factor in whether recuperating patients go home with healthcare system support, get transferred to an intermediate care facility, or stay in the hospital for a longer period of time.
Given this list of reasons that some patients stay longer in hospital, what can the SCOAP community do with length of stay (LOS) data? First, we can all recognize how we compare and benchmark off of one another. There is significant variation in LOS across SCOAP hospitals for elective colon and bariatric surgery, with some hospitals having unusually long LOS – did surgeons at these hospitals know this before SCOAP? Probably not, and it’s hard to address things that you don’t measure. Second, SCOAP hospitals have been effectively driving down LOS: first year SCOAP hospitals have a ½ day shorter LOS for colon resections than more recently joining SCOAP hospitals.
How have SCOAP hospitals driven down LOS? We found that reoperative complications were strongly linked to LOS. In SCOAP hospitals, the length of stay when there were reoperative complications was 19.9 days vs 7.4 when there were no complications. As reoperative complications have decreased, so too has LOS. Avoiding complications that require reoperation is probably the best way we can decrease LOS, and through process standardization (testing anastomoses, protective stomas, etc) that is a main purpose of SCOAP.
What esle can we do?
Surgeons are often frustrated by LOS as a metric of quality, because so much goes into LOS that’s non-medical. They feel they have little control over this metric. SCOAP has not been explicitly measuring non-medical reasons for LOS, but this may be something SCOAP surgeons start to dictate into their notes in the future so that it can be consistently reported. In June, SCOAP is adding additional measures on when diet was offered to patients to assess ileus, will be further evaluating discharge to intermediate care facility as an event of interest, and the SCOAP analytic team has been considering how to evaluate a patient’s geographic proximity to the hospital.
Why is LOS a big deal when others monitor surgeons? Part of the interest in LOS is that some percentage of patients with prolonged LOS have had untoward events, and measuring of LOS is an easy way for non-clinician driven initiatives to “measure” quality. SCOAP needs to control and “own” the LOS metric before others do it to us. Another reason is that every additional day in a hospital exposes patients to the risk of hospital acquired infections, decreased mobility with a risk of VTE, and exposure to the risk of iatrogenic injury. Lastly, there are potential financial issues in prolonged LOS.
Part of the financial interest in LOS from a hospital’s perspective is real (part is not as easy to measure) and relates to how a hospital’s resources might be used if they were not being used on our patient’s extended LOS. LOS becomes a directly measureable issue of hospital revenue in the uncommon scenario when there is effectively a bed shortage and when there is a revenue producing pipeline that cannot access the hospital but for over-utilization by longer staying surgical patients. This scenario is only the case with one or 2 hospitals in the state, and not all the time at those centers. For most of the other hospitals that are paid for an episode of care though a DRG, they do probably lose some money by providing services during the additional days that are not directly compensated. This is especially true if the hospital is paying for lots of overtime that might be avoided if the census was lower. However, care in the last few days of a hospitalization tends not to be overly expensive in terms of resources provided. It gets more complicated, however. Another consideration is that if patients are sent home while complications are in development and they are rehospitalized at another hospital, this may have some unintended consequences of delayed diagnosis, advanced presentations, and more expensive complication management. In fact, several payers are moving to not pay for rehospitalizations related to index procedures (see this article in a recent edition of The New York Times).
Beyond reducing complications-where else is there opportunity for improvement? Potentially modifiable components of care for reducing LOS in elective colon surgery include fast-track recovery with use of epidurals, Entereg use to avoid ileus, and laparoscopy. While it’s true that SCOAP hospitals have driven down LOS for colon surgery, why is it still so long compared to what we hear about fast-track surgery in Europe and some US centers? In a recent RCT of fast track surgery at four Swiss hospitals (Muller et al Gatroenterology 2009;136(3):842-7) this multimodal approach (including standard use of epidural catheters, steroids, volume restriction, expectation of early discharge, and early feeding/ambulation) resulted in significantly shorter LOS (median of 5 vs 9 days) and a lower rate of complications. Interestingly, though we don’t know who is using “fast track” protocols in SCOAP, when our analytic team looked at epidural use in elective colon surgery (24% of cases got one), we found that after adjusting for age, sex, insurance status, and comorbidity, those who did get an epidural actually had a longer length of stay (8.8 vs 7.7 days). Expect future reports looking at pain control to find out why it is that the epidural, often viewed as the key to quicker recovery, is not having an impact on LOS within our collaborative.
So what decreases LOS after colon surgery? The SCOAP analytic team also looked at laparoscopic techniques in elective colon surgery (33% of cases were laparoscopic) and found that after adjusting for these variables those who had a laparoscopic colon resection did have shorter LOS (6.6 vs 8.8 days). Lastly, some of you are starting to use Entereg, an opioid receptor blocker that recently came on the market to prevent ileus, and though it’s really too early to be certain about its impact, we’re not seeing differences in LOS yet. Stay tuned and ileus will be a feature in an upcoming SCOAP Box.
This may be more then what you wanted to know about LOS, but watch this one – it’s an important metric that we need to “own” if we want to improve care received and stave off top-down initiatives that are less well-informed than ours. In the meantime, SCOAP will continue to track on LOS and the factors that seem to be involved. It may or may not be important financially to hospitals, but at least in a subgroup of patients, longer LOS is a bad thing because it means something went wrong with the operation. These are complex issues, and the real issue here is that nothing gets better if we don’t monitor it. We look forward to your thoughts on LOS and how you as a collaborative want to work on LOS in the future.
How will we know if SCOAP is working?
Each month in The SCOAP Box, we show a different process of care or outcome metric getting better with each quarter of SCOAP participation. We also show the persistent variability in performance across the SCOAP hospitals and suggest there’s “more work to be done.” This month we highlight the dramatic improvements in rates of reoperative complications after elective colon and rectal resections. SCOAP hospitals have nearly halved the rate of this devastating complication using a combination of changes in patient selection and process of care that we will be highlighting in future issues. We have shown the rates of negative appendectomy reducing dramatically with benchmarking of diagnostic testing and length of stay after surgery being reduced by more than a day with increased used of care pathways. Most would agree that for our patients, our hospitals, and our own drive for excellence as surgeons, significantly reducing the rate of anastomotic leak, negative appendectomy, and length of stay are great achievements. This is something we should all be proud of and great motivation to tackle the things that have yet to improve in SCOAP.
As natural skeptics, many of us (including me) ask questions like: Are these real effects? Are these changes durable? Will they last once we shift focus? Is this just “regression to the mean” (outliers tend to perform like the average if monitored long enough) or a Hawthorne effect (things that you look at always get better just for looking at them)? To all these questions I think the answer is a qualified “maybe.” We will explore each of these as SCOAP moves on; we will continue to spot check performance as the spotlight of SCOAP shifts to different areas, and we will try to really link changes in process of care to outcome so we do not waste our time chasing metrics that have little impact on our patients. Distinguishing these effects from other changes going on at your hospital may always be a problem, but I would argue there is not that much general QI activity around anastomotic leak and negative appendectomy – to name only two. It may be that we are all just getting better over time, to which I say that at least SCOAP is able to prove that to everyone else. We are looking at temporal trend analyses and the effect of other interventions, but while all that is going on I would encourage the SCOAP community to not make these outcomes improvements (important as they may be) the only measures of SCOAP’s success.
So how will we know that SCOAP is working?
I propose we will know that SCOAP is working when we have recruited hospitals into the program and they have joined not because they HAVE to (because an insurer or Medicare forced them) but because administrators see that clinician ownership of surgical performance is better than administrator-forced performance improvement. Using this metric, we are well on our way – at 51 hospitals and growing and 85% of the state’s surgical care covered, we have moved beyond the tipping point and we are in the mode of picking up the late adopters. Already the question being asked by insurers, policy makers, doctors, media, and patients is, “why would this hospital NOT join SCOAP?” and to that end we can already claim success.
We will also know that SCOAP is working when hospital don’t just join SCOAP but when hospitals engage SCOAP into their strategic planning and execution, when they see SCOAP metrics as accountable, institutionalized, and “owned” by administrative/clinical teams, when they see SCOAP as a “staying in business” activity, and when they see the modest fees of SCOAP as an incredible bargain. To this end, SCOAP is just starting to achieve some success. With case studies at four SCOAP hospitals where we have engaged financial and administrative teams, we have broadened SCOAP’s focus beyond quality measures to also focus on standardization of care and supply choices that really impact measureable bottom lines and are showing the business case for SCOAP that measures in millions of dollars saved. We believe work in this area over the next year will help all SCOAP hospitals change from “joined” mode to “engaged” mode and that will be a real measure of success.
We will know that SCOAP is working when other surgeon and non-surgeon specialties reach out to the general surgical community and ask, “how can we ‘SCOAP’ our procedures?” This year, vascular surgeons from around the state began a grassroots initiative to build Vascular SCOAP (V-SCOAP), using the same surveillance, benchmarking, and intervention techniques of SCOAP to improve care delivery in vascular and endovascular procedures. This first-of-its-kind initiative will bring cardiologists, interventional radiologists, and vascular surgeons together to address critical issues in the field. I am confident that V-SCOAP will succeed and should launch this year. We also have similar energy emerging from the urologic community, and interest from plastics surgeons, orthopedic surgeons, and obstetricians to name just a few. Whether or not these other clinical areas are ready to be “SCOAPed” will depend on the leadership within those communities, if there is enough spirit of collaboration over competition to allow it, and the ability to present a compelling case to those working in the trenches and those paying the bills that SCOAP is worth it, but their interest alone is a sign that SCOAP is working. Their interest reflects what SCOAP surgeons already know – someone can do this to us or we can do it ourselves. SCOAP surgeons and hospitals have taken the proactive approach to having clinicians drive performance, and as that model expands, that is a tremendous measure of success.
We will know that SCOAP is working when we can move beyond the issues of reducing variability in process of care and outcome and start taking on the really difficult issues of appropriateness of care. Most of us are familiar with the Dartmouth Atlas showing widespread variation in the amount of care provided per person in different parts of the state and the country. Articles like those found in the June 1, 2009, issue of The New Yorker, “The Cost Conundrum” by friend of SCOAP Atul Gawande show that physician choices about when a procedure is indicated drive this variability in use and are a major component of our skyrocketing healthcare costs. When SCOAP is able to start a data-driven conversation and then impact things like cholecystectomy among patients with a single episode of biliary colic, colon resection after a single episode of uncomplicated diverticulitis, placement of a stent in the arteries of the lower leg in a patient with simple claudication it will really be a tremendous measure of success. We are not there yet, but we must get there eventually. The way SCOAP is doing this is by pulling together real experts in narrow areas of care, like all the surgeons who do rectal cancer surgery, to engage focused groups of surgeons in a conversation about how to define inappropriate care in our communities, how to reduce the inappropriate use of procedures, and how to get the patient’s voice heard in discussions about appropriateness of care. These groups are starting to ask these questions recognizing that surgeons and patients have an intimate ethics-bound relationship and that different decision making thresholds don’t necessarily reflect judgments about quality. You can bet insurers are interested in making these choices for us. This week in the other Washington they are looking for ways to make this happen with a pencil, but I would argue that as surgeons we can do this in a better way – in a way that will honor our profession and our patients. When SCOAP can tackle appropriateness, it will be a real sign that we are working to accomplish its full potential.
Colleagues, I think you get the idea. Many of us already think SCOAP is “working” in some very important ways – saving lives, improving care, and reducing costs. We have much to be proud of but much hard work ahead to really show that “working” translates into reaching our full potential. Because so many of you are committed to SCOAP and its principles, I am confident we’ll get there. I look forward to your continued questions, your partnership and our achievements as a collaborative.
We were asked by colleagues at Legacy Good Samaritan in Portland, “Why isn’t everyone doing a leak test to reduce serious complications after colon surgery?”
So why would SCOAP have to chase this metric? Why isn’t it 100% and what is it going to take to get us there? When we started SCOAP testing of left-sided colon, anastomosis was hovering around 50%, although some of this may have been under-reported in operative notes – but given the “billability” of the scope, it is hard to say. Here is what we heard when we started: some surgeons said they don’t like to add the time to the case; some believed that stressing an anastomosis can weaken it and actually cause leaks; many said that in cases when there was a leak after surgery, the leak test had been normal and therefore the test did not prevent a leak. And always the question was, while it seems like common sense, where is the evidence that leak testing makes a difference?
The best evidence for leak testing comes from a small RCT1 in the UK that randomized 145 patients receiving a colorectal anastomosis to “test” or “no test” sigmoidoscopic insufflation of air. A water-soluble contrast enema was performed on the tenth postoperative day. The two groups were well matched for age, sex, diagnosis, and other features. Eighteen (25%) air leaks were detected and repaired in the “test” group. After operation there were three (4%) clinical leaks in the “test” group and ten (14%) in the “no test” group (p = 0.04). There were eight (11%) radiological leaks in the “test” group and 20 (29%) in the “no test” group (p = 0.006). This was a small study at one center, and the very high rate of air leak in the OR (25%) had people scratching their head about whether this was generally applicable. In last month’s Archives of Surgery2, surgeons from the Lahey clinic report their series of over 2500 colorectal cases and focused on the 998 left-sided anastomoses, of which 82.7% were “leak tested”. Their findings were remarkably similar to the RCT findings, except the leak test was positive in only 7.9%. The recentness of the data and the much larger series should really help drive the use of leak testing. Overall they had a clinical leak rate of 4.8%. When patients were leak tested and the test was negative, the rate of subsequent leak was the lowest at 3.8%. If there were tests, and the test was positive (and the leak repaired), the risk of a postoperative leak was higher at 7.7%. Interestingly, the small group who were NOT tested had the highest leak rate of 8.1%. Testing seemed most important in patients with a circular-stapled anastomosis. Among patients who had circular-stapled anastomoses, the leak rate was low when testing was done (3.6% in those with negative tests, 4.8% in those with positive tests), and the rate was 21% among the small group of patients who did not have a leak test.
This is how it plays out in SCOAP: in the last 3 years, approximately 80% of left-sided SCOAP colons were tested – sometimes because of a routine checking practice and sometimes because of concern for a leak. When testing was done routinely (>90% of the time), the reoperative complication rate was 3.1% in those tested and 10% in those not tested. These rates are very similar to the Lahey clinic study and similarly show more than a doubling of the rate of leak when no testing was performed. SCOAP is building the evidence around leak testing and other common sense metrics. In the future, expect more detail about leak avoidance based on anastomosis type, the use of seam guards, and laparoscopic techniques.
Leak testing is a classic example of a common sense metric that SCOAP incorporated from the start and only now is more rigorous evidence developing to support its broader use. Because SCOAP has been ahead of the curve, we have already increased the use of testing from 55% to the 90% level among SCOAP hospitals. And this has meant great things for our patients. As would be expected from the recent study, the rates of reoperative complications across the state have improved with SCOAP’s work (see figures in July edition of “The SCOAP Box”), and while it is always difficult to link process of care changes to outcome, I would view leak testing as one of the many things we are working on that together appear to be improving the care our patients receive.
How is SCOAP driving changes in surgeon behavior (like increasing testing for leaks) to improve colon surgery? Before SCOAP most surgeons had never received a benchmarking report that showed them how they compared to their colleagues in the way they deliver care and the outcomes of that care. The power of benchmarking in driving behavior change is impressive. As surgeons, when we have not met SCOAP’s “top 10% benchmark”3 every one of us looks at how we differ in other ways, like testing for leaks, use of protective stomas, and avoiding elective surgery in patients with very low albumin (the last two will be addressed in upcoming SCOAP Box newsletters). Many of us have used this benchmarking to change behavior about things like leak testing so that we can say “we did everything we could” to avoid leaks after surgery. SCOAP has also raised this issue of leak testing in local, regional, and statewide meetings as a way for surgeons to take charge of the issue of avoiding leaks. Many of us were not trained in routine testing and are more inclined to practice the way we were trained. Creating opportunities to educate each other about different ways to do things and encourage the use of best practices is the heart of SCOAP.
With leak testing, you see the strength of SCOAP-surgeons defining, tracking, and encouraging best practices – actually driving better patient care. We are not at 100% testing yet, and SCOAP will continue to encourage more testing. SCOAP is also working to generate the evidence needed to encourage other process of care measures that improve outcome and avoid practices that do not help. Thank you for being a part of that process and making SCOAP successful.
We were recently asked by colleagues at Providence Everett, “What is the plan for other procedures in SCOAP?”
The Advisory Board has defined standards for SCOAP hospital participation. Full SCOAP participation requires 100% enrollment of patients in at least 2 modules (e.g., colorectal and vascular) and 10% auditing for other modules (e.g., appendectomy or bariatric) once a majority of benchmarks in those areas has been achieved. This allows SCOAP to be a “roving spotlight” that shifts its focus after a hospital achieves success but always includes a basic level of assessment so that the public can be assured that a SCOAP hospital is meeting the standards that its surgeons set. As other modules come online, hospitals can decide where they want to focus their data collection or can continue full enrollment in all modules.
So how should SCOAP decide where to shift the spotlight? Although we’ve often asserted that every procedure should be “SCOAP’d,” surgeons have to be mindful of which targets are most worthy of pursuing when developing a module. An important aspect of prioritizing expansion involves the expected benefits of SCOAP focusing on a target versus the costs of gathering data on that target. Procedures with frequent adverse outcomes, high variability in outcomes between sites, high procedural volume and simple in-hospital outcomes are highest priority targets, because addressing these would be very beneficial and data gathering would be straightforward. Lower priority SCOAP cases are those that are hard to track and include procedures where there appear to be only occasional adverse events, where there appears to be less variability in important outcomes, or when the procedure is simply not performed very often. Other important issues in setting SCOAP’s priorities include the readiness of the clinical community to work together in benchmarking and intervention. SCOAP is a voluntary collaborative activity, and clinicians have to lead the charge for this to work. Another consideration has to do with the operational aspects of outcome assessment. It is easiest to capture outcomes when they occur in the hospital and where there is little subjectivity in defining outcomes. While the things that happen to patients once they are at home are important, these outcomes often require addressing subjective issues and a different data-gathering platform with unique challenges.
SCOAP’s immediate plan for expansion is to finalize the modules for pediatric and vascular surgery, both of which are high variability specialties that have relatively high volume and meet all the criteria for high-priority. The next target will be a high-risk/low-risk module set for a Winter 2009/Spring 2010 launch, and we are currently assessing the feasibility and readiness of developing Urologic SCOAP for next summer.
High-risk/Low-risk SCOAP is an interesting venture that should impact every SCOAP hospital. High-risk SCOAP will cover esophageal, pancreatic, and liver resections. While these procedures are performed relatively infrequently and are concentrated at less than 10 hospitals, there is important variation in outcome that makes them excellent SCOAP targets (see August 2009 SCOAP Box). There is also little sign that things are getting better on their own ((see August 2009 SCOAP Box). To create high-risk SCOAP, we are calling on all surgeons in the state who are involved in these procedures to join together in specialty community groups that will participate in focus groups and web/phone-based conversations to create the dataset that will be of highest value in benchmarking and intervening to improve quality. When there are national data points or data sets, SCOAP always aligns with them, and where there are evidence-based metrics we always incorporate them. Unfortunately, as many of you know, there are few standards and established metrics in some of these arenas. SCOAP’s pragmatic, flexible, and evolving dataset generation technique will allow us to incorporate common sense metrics based on the measures derived from your considerable experiences with these procedures, best practices, and exploratory metrics that we can learn about through the process. Because high-risk SCOAP involves lower volume procedures, we expect that the data burden will be low but that the potential impact on the variability we are seeing will be dramatic. Stay tuned for more updates.
Lower-risk SCOAP is a completely different concept and will pose some unique challenges. Much of the fall and winter will go into an exploration of the feasibility of creating a module that focuses on hernia, gall bladder, endoscopy, and breast surgery. Most surgeons have told us that they like SCOAP benchmarking but that the vast majority of procedures they do now are ambulatory or short-stay procedures and that the adverse outcomes in these cases are infrequent and occur after discharge. SCOAP planning for any new module (including lower-risk procedures) starts by asking the question, “What outcomes are relevant to patients and potentially modifiable?” For ambulatory surgery, there are some procedure-specific issues – gall bladders and bile duct injury, surgical site infection and recurrence in hernia, hematoma in breast, and perforation in endoscopy – but these are very infrequent events, and SCOAP might have to track thousands of procedures to come up with enough cases that had the problem to meaningfully benchmark it. Alternatively, anesthetic-related issues – post-operative nausea and vomiting (PONV) and urinary retention – occur quite commonly, appear to be very modifiable through prophylaxis, and represent an area of serious concern for patients (PONV will be the focus of an upcoming SCOAP Box newsletter).
To address this issue of high volume, low risk procedures and the overwhelming data burden they pose, SCOAP is planning two approaches. The first changes the data collection scheme from a detailed dataset in 100% of the patients to a detailed data collection only in “cases” (and use of a case-control design) accompanied by a very limited “100% dataset” that allows for calculations of rates. The case-control approach identifies cases (e.g., patients with severe PONV, those readmitted for urinary retention, and those with a bile duct injury) and gathers detailed information about what happened in their operation as well as similar detailed information on a matched patient that did not have that outcome. This will help explain whether the SCOAP metrics adopted for PONV or bile duct injury were met in the cases as well as the controls and should help hospitals and surgeons understand where there are modifiable factors that can be addressed. How SCOAP captures post-discharge events is an area of considerable focus: a multi-tiered approach using data from re-hospitalizations at the index center, admissions within 30 days to other hospitals, and direct patient outreach will fill out the complete picture of what happens to our patients after discharge. SCOAP is developing the infrastructure to make this an automated (non-human data abstracting) process. This technology will gather patient-reported, automated results through use of the full spectrum of contacts (e-mail, text messaging, cell phone, 1-800 numbers, and letters) and plans to incorporate these data into the feedback of the lower-risk surgery module.
This is an ambitious project that aims to redefine the way patients report information to hospitals and surgeons. We see this as a great way to help SCOAP surgeons improve care by benchmarking outcomes of importance to patients. By including questions about quality of life, functional status and customer satisfaction, this will also be a valuable tool for improving service delivery and seeing your practice through the eyes of your patients. Stay tuned for updates about SCOAP expansion. In the meantime, we encourage you to get more involved. We need your voice to be a part of SCOAP planning in order to really fulfill our mission of being a surgeon-led collaborative.