Troponin Leak Postop – what does it mean?

vision studyTwenty years ago perioperative myocardial ischaemia was a relatively easy thing to diagnose – we checked ECG looking for ST segment and T wave changes, and looked for an MB-CK rise. Then troponin arrived, and suddenly the proportion of patients with perioperative ischaemia increased drastically. For many of us, the report of a “postoperative troponin leak” results in a shoulder shrug: we don’t know what it means, we don’t really know the long term implications.

Thankfully, a landmark study, VISION (click here), has provided us with quality epidemiologic data. This was a cohort study of 15000 patients >45 years that underwent non cardiac surgery and had troponin T (TnT) measured in the first 3 postoperative days. All patients had to have procedures that required overnight stay in hospital. The main outcome measure was 30 day mortality.

After 30 days 1.9% of patients had died. Patients were more likely to die if their peak TnT level was 0.02 ng/ml (versus reference range of <0.01 ng/ml). This occurred in 11.6% of patients. The greater the TnT level, the more likely the patient was to die. They were able to stratify risk depending on TnT levels. Patients with a peak TnT value of 0.01 ng/mL or less, 0.02, 0.03-0.29, and 0.30 or greater had 30-day mortality rates of 1.0%, 4.0%, 9.3%, and 16.9%, respectively (figure above).

Risk was expressed in terms of Hazard Ratio (HR): greater HR = more likely adverse outcome with 1 being equivalent to no additional risk, <1 lower risk, >1 higher risk. Peak TnT of 0.02 ng/mL (adjusted hazard ratio [aHR], 2.41; 95% CI, 1.33-3.77); 0.03 to 0.29 ng/mL (aHR, 5.00; 95% CI, 3.72-6.76); and 0.30 ng/mL or greater (aHR, 10.48; 95% CI, 6.25-16.62).

Who was at increased risk? The older the patient the higher the risk. Emergency surgery, general surgery, neurosurgery were associated with increased risk. Vascular surgery was not, although the presence of peripheral vascular disease, COPD, previous stroke, coronary arterial disease and cancer did predict adverse outcome. Diabetes, obesity, afib, OSA, hypertension, orthopaedic/thoracic urology surgery – did not predict adverse events.

Conclusions: these data demonstrate the efficacy of TnT measurement in determining perioperative prognosis. 1in 25 patients with a peak TnT measurement of 0.02ng/mL,1 in11patients with a peak TnT measurement of 0.03 to 0.29ng/mL, and 1 in 6 patients with a peak TnT measurement of at least 0.30ng/mL will die within 30 days of surgery. Two questions arise from this study: 1. should we be routinely measuring TnT postoperatively in surgical inpatients >45 years; 2. If the patient has a troponin leak – what should be do then: PCI, aspirin, clopidogrel, statins, betablockers, all of the above, none of them? Will routine measurement of TnT result in a dramatic increase in cardiology consultations with little evidence that there are interventions that may improve outcomes in this setting?

EUSOS follow up – is it the beds?

Over the next few months I am sure that the real reasons for the comparatively poor outcomes of Irish patients in the EUSOS study will emerge. In the meantime, we can only guess the reasons. Aside from blaming surgeons for poor patient selection (which is suspiciously convenient), case volume may be a problem, the time of day (exhaustion), the amount of emergency surgery (including case volume) or the issue may lie in our own backyard – in the availability of beds for high risk postoperative patients. Emergency surgery patients, in particular, do poorly.

A US study of 25,710 nonemergency colorectal resections performed at 142 hospitals reported a 1.9% (492 patients) mortality rate. For emergency colorectal resection the mortality rate was 15.3% (780 of 5,083 patients). Fifty percent of emergency surgery patients had at least 1 complication versus 24% of elective surgery patients. This is horrifying.

The first report of the UK emergency laparotomy network (here), published in the BJA, presents similar mortality data. As a guide, mortality rates for major elective general surgery have been reported as follows: colorectal resection – 2.7%,  oesophagectomy – 3.1%, gastrectomy – 4.2% and liver metastasis resection – 1%. In this study (data from 1853 patients were collected from 35 NHS hospitals) the unadjusted 30 day mortality was 14.9% for all patients and 24.4% in patients aged 80 or over.

We are aware that emergency surgery patients come in at all hours of the night and are frequently operated on by junior doctors. The time of day was an issue (table below) – 30 day mortality was 50% higher if surgery took place between midnight and 8am. Obviously confounders may be present – surgeons may only take the sickest patients to theatre at night, and this may represent selection bias.

Time of day* n Consultant anaesthetist present (%) Consultant surgeon present (%) 30 day mortality (%)
08:00–17:59 1044 75.2 80.8 14.2
18:00–23:59 442 54.8 67.7 17.8
00:00–07:59 152 40.8 61.8 20.3

Bad outcomes occurred for patients admitted under a medical service who actually had a surgical problem, increasing age, increasing ASA physical status.

What about beds? “Of the patients who were felt to need intensive care immediately after surgery, 99% were transferred to a level 3 bed. Similarly, 89% of those who were judged to require a high-dependency bed received this level of care, with a further 4% receiving level 2 care in an ICU bed. Mortality in patients returning to the ward (level 1) was 6.7%, HDU 10.1%, and ICU 30.7%. 2.2% of patients were cared for after operation in an extended recovery area (presumably because there was no HDU bed available), and this group had a mortality of 13.5%. For the group of patients aged 60 or greater, and of ASA III or more (∼50% of all patients), 22% returned to the general ward after operation and had a mortality of 17.8%.” One must presume that this 22% represented at least part of the 11% that didn’t get the needed HDU beds. Hence, one could crudely argue that the patients that needed HDU beds but didn’t get them had an absolute mortality risk increase of 7.7% (the authors do not give us sufficient data to make direct comparisons, but more than 50% of patients were >60y and ASA III or greater). The overall mortality for patients sent to a regular ward was 6.7%, which appears to be very high when compared with data from general elective surgery (above). However, a recent study of all 160,920 patients who underwent bowel resection for colorectal cancer between 1998 and 2006 in the English NHS reported a mortality rate of 6.7%

These data at least suggest that lack of availability of a HDU/ICU bed significantly increases the risk of poor postoperative outcomes for emergency surgical patients.

The utilization of critical care services has been known to be suboptimal for many years. A previous study, published in Anaesthesia (here) looked at 26000 patients undergoing surgery in an NHS trust: “only 852 (35.3%) high-risk patients were admitted to a critical care unit at any stage after surgery. Of 294 high-risk patients who died, only 144 (49.0%) were admitted to a critical care unit at any time and only 75 (25.6%) of these deaths occurred within a critical care area. Mortality rates were high amongst patients discharged and readmitted to critical care (37.7%) and amongst those admitted to critical care following initial postoperative care on a standard ward (29.9%).” So, inadequate numbers of ICU/HDU beds are associated with poor outcomes, and early discharge (presumably for bed pressure) and readmission is associated with 1/3 of patients dying.

Ireland has a similar number of critical care beds per 100,000 population (6.5/100,000) to the UK (6.6/100,000). In a recent pan European study conducted by Andy Rhodes (here), Ireland ranked 26th out of 31 (UK was 25th) in critical care bed numbers per 100,000. The European average was 11.5. Overall, Ireland ranked 28th/31 for number of acute care beds and  23rd out of 31 for ICU beds as a % of acute care beds. So, we have very few beds for sick patients, and of these very few of them are critical care beds. Ireland spends 7.2% of GDP on healthcare (15th/31) and has the 6th highest GDP in proportion to ICU beds. In other words – we spend very little money comparatively on critical care compared with Europe. This might reflect the fact that we have the 2nd youngest population in Europe (10.4% are 65 or older).

In summary – is lack of critical care beds a likely factor for Irelands poor showing in EUSOS: almost certainly. Do these studies fully explain the difference – no! Unfortunately, the OR death was still 2.6 times the UK with a similar number of ICU/HDU beds. It could be argued that the bed numbers are inflated in Ireland, due to poor distribution between hospitals – community hospitals have underused ICU beds, referral centers have inadequate capacity. But that is another discussion….

At last – Chloride is nephrotoxic

For years I have been trotting around the world telling everyone that NaCl 0.9% is evil, because each litre delivers 50mmol of HCL and chloride is nephrotoxic. This belief has come from a series of studies in volunteers (reduced GFR, reduced splanchnic perfusion, reduced cortical blood flow) and observations (increased contrast nephropathy with NaCl versus NaHCO3. I suggested that the CHEST trial failed to prove that HES was dangerous because the control fluid was saline. But where was the real proof of nephrotoxicity.
Here it is in JAMA (click here).

A group in Melbourne, Australia, performed a sequential patient cohort study during 2 time periods: in phase 1 any IV fluid could be used; in phase 2 (the following year), chloride rich fluids were unavailable, so balanced salt solutions only could be prescribed.

Chloride administration fell considerably: from 694 to 496 mmol/patient from the control period to the intervention period. Patients in the chloride rich period had significantly worse renal outcomes: the mean serum creatinine level increase while in the ICU was 22.6 μmol/L (95% CI, 17.5-27.7 μmol/L) vs 14.8 μmol/L (95% CI, 9.8-19.9 μmol/L) (P = .03), the incidence of injury and failure class of RIFLE-defined AKI was 14% (95% CI, 11%-16%; n = 105) vs 8.4% (95% CI, 6.4%-10%; n = 65) (P <.001), and the use of RRT was 10% (95% CI, 8.1%-12%; n = 78) vs 6.3% (95% CI, 4.6%-8.1%; n = 49) (P = .005). In other words – patients given balanced chloride fluids had a 3.7% reduction in the risk of needing dialysis (NNT <30). As you would expect, there was no difference in mortality figures.

The accompanying editorial can be read here.

No I won’t do it and here is the proof!

As a junior doctor how many times were you called to replace an iv catheter on a veinless patient because with was 3 days old (and “hospital policy” and all that). There was no point asking to see the evidence on which this “policy” was based. Whatever! – here is the counter evidence, and it is in the Lancet (here).

The study in question was a multicentre, randomised, non-blinded equivalence trial recruited adults (≥18 years) with an intravenous catheter of expected use longer than 4 days from three hospitals in Queensland, Australia in 2008-09 (why so long to publish?). There were 3283 patients randomised (5907 catheters- 1593 clinically indicated; 1690 routine replacement).

The mean time the iv cannulae lasted when they were in situ on day 3 was 99 h (SD 54) when replaced as clinically indicated and 70 h (13) when routinely replaced. In other words – not routinely changing the catheter resulted in it being in place for 1.25 extra days. Phlebitis occurred in 114 of 1593 (7%) patients in the clinically indicated group and in 114 of 1690 (7%) patients in the routine replacement group: ABSOLUTELY NO DIFFERENCE, NONE, STOP ASKING ME LEAVE ME ALONE!

So, if the iv site looks ok – it is ok. Don’t go prodding the patient.

Just when you thought it was unsafe….HES again!

A couple of weeks ago I announced the imminent death of colloid. Now it’s back with another “Safe” trial (known as CHEST) from our colleagues in Australia and New Zealand (here). The study enrolled a colossal number of patients (7000) to either isotonic saline (IS) or Voluven (R). This is a 130/0.4 tetrastarch in isotonic saline. I now understand the kerfuffle over the 6s trial (of tetraspan) that mislabeled 130.42 in balanced salt as 130/0.4 (read here). The paper should be subtitled: “ours – the starch that doesn’t kill!” Perhaps…
The authors randomly assigned 7000 patients who had been admitted to an intensive care unit (ICU) in a 1:1 ratio to receive either 6% HES with a molecular weight of 130 kD and a molar substitution ratio of 0.4 (130/0.4, Voluven) in 0.9% sodium chloride or 0.9% sodium chloride (saline) for all fluid resuscitation until ICU discharge, death, or 90 days after randomization. The patients were permitted 50ml/kg HES per day and then were given saline. Similar to the SAFE (albumin) trial, clinicians were permitted to resuscitate patients according to their own goals and preferences.
The primary outcome was death within 90 days. Secondary outcomes included acute kidney injury and failure and treatment with renal-replacement therapy.
There was no mortality difference. A total of 597 of 3315 patients (18.0%) in the HES group and 566 of 3336 (17.0%) in the saline group died (relative risk in the HES group, 1.06; 95% confidence interval [CI], 0.96 to 1.18; P=0.26). There was no significant difference in mortality in six predefined subgroups. Renal-replacement therapy was used in 235 of 3352 patients (7.0%) in the HES group and 196 of 3375 (5.8%) in the saline group (relative risk, 1.21; 95% CI, 1.00 to 1.45; P=0.04).
HES was associated with less renal injury than saline, by RIFLE criteria, but post hoc creatinine and urinary output were worse for HES In the HES and saline groups, renal injury occurred in 34.6% and 38.0% of patients, respectively (P=0.005), and renal failure occurred in 10.4% and 9.2% of patients, respectively (P=0.12). There was a 1.2% absolute increase in the risk of needing renal replacement therapy (p<;0.5) in the HE’S group.
HES was associated with significantly more adverse events (5.3% vs. 2.8%, P<;0.001). These included itching, skin rash and “other” (not explained).

A few comments: the mortality rate for a critical care study was astonishingly low, suggesting that the addition of surgical patients (42%) and some entry restrictions may have biased the study [Of the 7000 patients 2,876 were admitted from the operating room]. Almost 10% of patients came from another hospital. Only 1 in 4 came from the emergency department. So there is likely lead-time bias (incidentally, this distribution is near identical to the 6S study. It is highly unlikely that the majority of patients received, exclusively, one of these investigation fluids prior to ICU admission. Patients were in the ICU for 10 or 11 hours prior to randomization: the “golden window”. The patients should have been resuscitated by this stage. This is suggested by the almost ludicrously small amount of fluid that patients received day 0 (see below).

As expected HES patients received less fluid early on, but this did not translate into better outcomes. What is surprising is how little fluid the patients received in the first 24 hours (1500ml to <;2000ml net fluid balance). Patients received between 1000ml (HES) and 1500ml (IS) in addition to study fluids. In fact they seem to have gotten a lot more non study fluid than study fluid. This is presumably due to the 50ml/kg limit (400ml for an 80kg patient). As the on study fluid of choice was IS, this was really a crystalloid plus colloid versus crystalloid study. Indeed patients almost exclusively received IS
It is remarkable how little fluid the patients accumulated over the first 3 days (by the end of day 3 the IS group appear to have a net negative balance). What ever way you look at it, patients received significantly less fluid than in the 6S study (nearly 6L day 1). I return to my previous observations: nearly 50% surgical patients, with resolving stress responses, patients probably already resuscitated before randomized to the study.

It would have useful to know the electrolyte and acid base status. These patients all received a lot of chloride: what proportion of the had hyperchloremic acidosis?

It is very hard to make anything of the renal function tests in this study. On first sight the 36% rate of renal dysfunction at baseline was similar to 6S. But Scandanavian patients were significantly sicker. They had a mortality rate of 43-50%, consistent with other sepsis trials (such as VISEP). RRT use in the 6S study was 16% in the Ringer’s acetate group versus 5.8% in the IS group in this study. So I would be inclined to ignore RIFLE numbers and consider “real” kidney injury to be represented by the need for RRT.

So, how to evaluate this study? Does this study demonstrate the safety of HES in critical illness? No, it just shows that HES doesn’t increase mortality versus isotonic saline. They may be equally bad. Does this paper conflict with the 6S study? Only 30% of patients in the study were septic, the mortality was substantially lower in this study and it is likely that HES worsens outcomes in patients that require more of it (i.e. sicker patients). The onus of proof is on the intervention: HES demonstrates no mortality advantage over crystalloid, it may worsen outcome, it may be associated with more organ (particularly kidney) dysfunction. I am less likely to use these products in ICU. However, there is still a small argument for colloid administration in the peri operative period based on a series of oesophageal Doppler studies (from the UK). It is highly unlikely that 500ml HE’S will harm a patient. However, I don’t quite see the point: HES is expensive compared with crystalloid and only appears to have a marginally better volume expanding effect: why take the risk without clear benefit.
I presume some clever scientist will take this, the 6S and other crystalloid-colloid studies and inform us about all of the residual questions in a nice meta-analysis. In the near future, it would be helpful if the CHEST investigators would give us data on chloride levels and acid-base status. Perhaps their next study should be to compare isotonic saline to balanced salt solutions.
Finally, congrats again to the ANZICS trial group for showing the rest of us what can be done.

Regional or General Anaesthhesia for Hip Fracture

A 78 year old female patient is brought to the operating room with a fractured hip. She tells you that she wants to “go asleep” for the operation. How do you advise her?
Two papers in July’s Anesthesiology have shed light on this issue. Both studies mine large databases and so care must be taken to avoid over interpretation of data.
Neuman and colleagues (read here) looked at data at 126 New York hospitals over 2 years. Surprisingly, of 18,158 patients only 5,254 (29%) received regional (neuraxial) anesthesia. One in 40 patients died in hospital and, unadjusted, there was no difference in the rates of mortality between GA and RA. Patients receiving regional anesthesia experienced fewer pulmonary complications (359 [6.8%] vs. 1,040 [8.1%], P <0.005). Regional anesthesia was associated with a lower adjusted odds of mortality (odds ratio: 0.710, 95% CI 0.541, 0.932, P = 0.014) and pulmonary complications (odds ratio: 0.752, 95% CI 0.637, 0.887, P<0.001). The benefits associated with regional anesthesia accrued to patients only with intertrochanteric fractures; regional did not benefit patients with femoral neck fractures.

Memtsoudis and colleagues (read here) mined a 530,000 national (USA) database of patients undergoing primary hip and knee arthroplasty. One in 30 patients utilized critical care services. Patients that underwent general anaesthesia, elderly patients and those that has cardiopulmonary complications, were significantly more likely to use critical care. As one would expect, admission to the ICU was associated with significantly increased mortality (2.5% versus 0.1%). Patients were also more likely to enter ICU if they were in smaller non teaching hospitals and if they had hip rather than knee surgery.

Interestingly, this study utilized the Deyo index (here) rather than ASA physical status score. The Deyo index appears to be a strong predictor of outcomes in patients having major orthopedic surgery (here). Co-morbidlty indexes are very useful in clinical practice to predict risk (here). Indeed, the Deyo index and ASA-PS score have been used together to demonstrate adverse outcomes (here).

EuSOS study published – and it’s not pretty!

46,539 patients from all over Europe were recruited to the The European Surgical Outcomes Study over 7 days in April 2011 (read here). Day cases, cardiac and neurosurgical patients were excluded. The overall mortality rate was 4% (nearly 1 in 20 patients). 8% of patients were admitted to ICU or HDU at some stage – but, astonishingly, 73% of those who died never saw a critical care practitioner.
For Ireland 856 patients were recruited into the study; 66 went to critical care beds postoperatively. Median hospital stay was 3 days (1.0-6.0). 6.4% died in hospital, with an unadjusted (for severity of illness) odds ratio of death (compared with the UK) of 1.86. When severity of illness was taken into account the OR of death was 2.61. This puts us down the scale of outcomes with Croatia, Slovakia (better), and Romania and Latvia (marginally worse).
What is truely frightening about these data – is that the reference country, the UK, aside from having a similar population to ours, had worse outcomes than they had expected (mortality 3.6% rather than the predicted 1.6%).
It could be argued that these data are skewed by relatively low numbers, recruitment exclusively in academic medical centers (private hospitals cherry pick the healthiest elective surgery patients), the significant limitations of the ASA physical status grade (between 2 and 3 there really should be 3 more grades – clinicians may have also reported patients as a ASA-PS 2 when they really were a 3), reporting bias etc. Alternatively, our patients might do badly because of  weaker nursing care at ward level and fewer critical care beds per head of population.
If the anaesthesia and critical care community in Ireland wants to look into this further, perhaps a worthwhile study would be an enthusiatic clinician to pull out the charts of all 856 patients and figure out why Ireland did so badly. Comments?

Hyperoxia and Surgical Site Infections: is oxygen beneficial?

Using high inspired concentrations of oxygen in the perioperative period may reduce the risk of surgical site infections for patients undergoing colo-rectal surgery. It does not appear to confer benefit for other patient groups.

We live side by side with an element that both feeds us and damages us simultaneously: oxygen. Reactive oxygen species cause lipid peroxidation of cell membranes and disrupt DNA. They interfere with gene expression and cause altered cell growth and necrosis. This happens all the time, and we have developed anti-oxidant scavenging systems for clearing up the debris. So oxygen is toxic. Conversely, oxygen kills bacteria – facilitating the activity of neutrophils, thus enhancing immune function, it is anti-inflammatory,1 it is a vasoconstrictor (may reverse vasoplegia) and it redistributes blood flow to the kidneys and splanchnic circulation.2-4 Oxygen is potentially therapeutic in sepsis.5

Surgical site infections (SSI) result in significant morbidity, delayed hospital discharge and increased healthcare costs. There is a known association between SSI and hypoperfusion, contaminated wounds, perioperative hyperglycaemia and hypothermia6 and obesity. It has long been proposed that the use of perioperative hyperoxia to high risk patients may result in a reduction in the risk of SSIs. The converse argument is that hyperoxia is toxic to the lungs7;8 and results in increased atelectasis and, potentially, an increase in postoperative pulmonary complications.9

The scientific rationale for preoperative hyperoxia is that oxidative killing by neutrophils, the primary defence against surgical pathogens, depends critically on tissue oxygenation.10 Hopf and colleagues11 performed a non interventional, prospective study of subcutaneous wound oxygen tension(PsqO2) and its relationship to the development of woundinfection in surgical patients. One hundred and thirty general surgical patients were enrolled and PsqO2 was measured perioperatively. There was an inverse relationship between wound oxygen tension and the risk of developing surgical site infections (SSI). They hypothesized that manipulating FiO2 may increase PsqO2 and reduce SSIs.

Grief et al12 randomly assigned 500 patients undergoing colorectal resection to receive 30 percent or 80 percent inspired oxygen during the operation and for two hours afterward. This was a very well constructed study. Anaesthetic treatment was standardized, and all patients received prophylactic antibiotic therapy, standardized fluid regimens and kept euthermic perioperatively. Wounds were evaluated daily until the patient was discharged and then at a clinic visit two weeks after surgery. The arterial and subcutaneous partial pressure of oxygen was significantly higher in the patients given 80 percent oxygen than in those given 30 percent oxygen. The duration of hospitalization was similar in the two groups. Among the 250 patients who received 80 percent oxygen, 13 (5.2 percent; 95 percent confidence interval, 2.4 to 8.0 percent) had surgical-wound infections, as compared with 28 of the 250 patients given 30 percent oxygen (11.2 percent; 95 percent confidence interval, 7.3 to 15.1 percent; P=0.01). The absolute difference between groups was 6.0 percent (95 percent confidence interval, 1.2 to 10.8 percent) NNT 15.

These data were confirmed by a smaller study from Spain. Belda et al13 undertook a double-blind, randomizedcontrolled trial of 300 patients aged 18 to 80 years who underwentelective colorectal surgery. Patients were randomly assigned to either30% or 80% fraction of inspired oxygen (FIO2), intraoperatively,and for 6 hours after surgery. Anaesthetic treatment and antibioticadministration were standardized.A total of 143 patients received 30% perioperativeoxygen and 148 received 80% perioperative oxygen. Surgical siteinfection occurred in 35 patients (24.4%) administered 30% FIO2and in 22 patients (14.9%) administered 80% FIO2 (P=.04). Therisk of SSI was 39% lower in the 80% FIO2 group (relative risk[RR], 0.61; 95% confidence interval [CI], 0.38-0.98) vs the30% FIO2 group. After adjustment for important covariates, theRR of infection in patients administered supplemental oxygenwas 0.46 (95% CI, 0.22-0.95; P = .04). Similar results were reported by Bickel and colleagues, in a 210 patients with acute appendicitis (5.6% versus 13.6%, p = 0.4, ARR 7 NNT – 13).14

Pryor et al claimed opposite results.15 This study included 165 patients that were undergoing general surgery, and were randomized to 30% or 80% oxygen. The overall incidence of SSI was18.1%. In an intention-to-treat analysis, the incidence of infectionwas significantly higher in the group receiving FIO2 of 0.80than in the group with FIO2 of 0.35 (25.0% vs 11.3%; P = .02).FIO2 remained a significant predictor of SSI (P = .03) in multivariateregression analysis. Patients who developed SSI had a significantlylonger length of hospitalization after surgery (mean [SD], 13.3[9.9] vs 6.0 [4.2] days; P<.001).

This study was criticized for a number of reasons. It is unclear whether or not the group assignment was truly blind. Tissue oxygenation was not blind. Wound infection was identified by retrospective chart review, a highly unreliable technique. There was no standardization of fluid therapy, temperature or antibiotic prophylaxis. Patients receiving80% oxygen were more likely to be obese, had longer operations,and lost more blood. All these factors may be associated withincreased risk of SSI. Significantly more patients in the high FiO2 group went back to the PACU intubated post op. Finally, the incidence of wound infections, at 25%, was high in the hyperoxic group compared with the study by Grief, 12 but similar to the control group in the study by Belda.13

Maragakis et al16 undertook a case-control retrospective review of SSIs in patients undergoing spinal surgery. Two hundred and eight charts were reviewed. The authors claimed that the use of an FiO2 of <50% significantly increased the risk of SSI (OR, 12; 94% CI, 4.5-33; P < 0.001). This study has the same flaws as that by Prior and colleagues,(50) albeit with opposite results.

Myles et al 17 enrolled a 2,050 patients into a study that randomized them to either FiO2 of 80% or 30%, plus 70% nitrous oxide. Patients that were given a high FiO2 had significantly lower rates of major complications (odds ratio, 0.71; 95% confidence interval, 0.56-0.89; P = 0.003) and severe nausea and vomiting (odds ratio, 0.40; 95% confidence interval, 0.31-0.51; P < 0.001). Among patients admitted to the intensive care unit postoperatively, those in the nitrous oxide-free group were more likely to be discharged from the unit on any given day than those in the nitrous oxide group (hazard ratio, 1.35; 95% confidence interval, 1.05-1.73; P = 0.02). It is unclear whether these data represent a beneficial effect of oxygen or a detrimental effect of nitrous oxide.

The Proxi trial 18 included 685 patients in 14 Danish hospitals. Patients were randomized to 80% versus 30% oxygen. Temperature, fluid therapy and type of surgery were not controlled. Similar to the Pryor trial, the incidence of SSIs were in excess of 20% (20.1%) in the control group, not significantly different from the study group (19.1%). There was no difference in pulmonary complications between the groups. Clearly the extraordinarily high number of SSIs in both groups made a statistically significant difference in outcomes unlikely. A large number of patients had undergone emergency surgery and had contaminated wounds. Hence, a direct comparison with previous studies cannot be made.

However, comparisons have been made and here have been several meta-analyses (MA) of hyperoxia and surgical site infections. These differ in outcomes depending on whether or not one includes the Myles17 data. Where Myles’s study is included, the MA supports hyperoxia.19 Where it is excluded – MAs routinely exclude papers for reasons that are not always obvious – hyperoxia is shown not to be beneficial.20 My own conclusion is that there is tremendous heterogenicity between these studies: well controlled studies of colonic surgery where anaesthesia and perioperative care was standardised resulted in better outcomes. Poorly controlled studies (Pryor / Meyhoff), without standardisation resulted in very high levels of SSI in both groups. The excess adverse outcomes in the Pryor study suggests that there were substantial differences between the groups in terms of type and length of surgery, severity of illness etc. and that this study was fatally flawed.

My conclusion: if you are providing anaesthesia for bowel surgery, and will not be using nitrous oxide, 80% oxygen is unlikely to be harmful, and is potentially beneficial. Whether or not to extend this hyperoxia into the postoperative period is very controversial.


1.    Nathan C: Oxygen and the inflammatory cell. Nature 2003; 17: 675-6

2.    Bitterman H, Brod V, Weiss G, Kushnir D, Bitterman N: Effects of oxygen on regional hemodynamics in hemorrhagic shock. Am J Physiol 1996; 40: H203-H211

3.    Cason BA, Wisneski J, Neese RA, Stanley WC, Hickey RF, Shnier CB, Gertz EW: Effects of high arterial oxygen tension on function, blood flow distribution, and metabolism in ischemic myocardium. Circulation 1992; 85: 828-38

4.    Plewes JL, Farhi LE: Peripheral circulatory responses to acute hyperoxia. Undersea Biomed Res 1983; 10: 123-9

5.    Bitterman H: Bench-to-bedside review: Oxygen as a drug. Critical Care 2009; 13: 205

6.    Kurz A, Sessler DI, Lenhardt R: Perioperative Normothermia to Reduce the Incidence of Surgical-Wound Infection and Shorten Hospitalization. New England Journal of Medicine 1996; 334: 1209-16

7.    Fisher AB: Oxygen therapy, side effects and toxicity. Am Rev Respir Dis 1980; 122: 61-9

8.    Bitterman N, Bitterman H: Oxygen toxicity. Handbook on Hyperbaric Medicine 2006; 731-66

9.    Hedenstierna G, Edmark L, Aherdan KK: Time to reconsider the pre-oxygenation during induction of anaesthesia. Minerva Anestesiol. 2000; 66: 293-6

10.    Overdyk FJ: Bridging the Gap to Reduce Surgical Site Infections. Anesthesia & Analgesia 2010; 111: 836-7

11.    Hopf HW, Hunt TK, West JM, Blomquist P, Goodson WH, III, Jensen JA, Jonsson K, Paty PB, Rabkin JM, Upton RA, von Smitten K, Whitney JD: Wound Tissue Oxygen Tension Predicts the Risk of Wound Infection in Surgical Patients. Archives of Surgery 1997; 132: 997-1004

12.    Greif R, Akca O, Horn EP, Kurz A, Sessler DI, The Outcomes Research Group: Supplemental Perioperative Oxygen to Reduce the Incidence of Surgical-Wound Infection. The New England Journal of Medicine 2000; 342: 161-7

13.    Belda FJ, Aguilera L, Garcia de la Asuncion J, Alberti J, Vicente R, Ferrandiz L, Rodriguez R, Company R, Sessler DI, Aguilar G, Botello SG, Orti R, for the Spanish Reduccion de la Tasa de Infeccion Quirurgica Group: Supplemental Perioperative Oxygen and the Risk of Surgical Wound Infection: A Randomized Controlled Trial. JAMA: The Journal of the American Medical Association 2005; 294: 2035-42

14.    Bickel A, Gurevits M, Vamos R, Ivry S, Eitan A: Perioperative Hyperoxygenation and Wound Site Infection Following Surgery for Acute Appendicitis: A Randomized, Prospective, Controlled Trial. Archives of Surgery 2011; 146: 464-70

15.    Pryor KO, Fahey TJ, III, Lien CA, Goldstein PA: Surgical Site Infection and the Routine Use of Perioperative Hyperoxia in a General Surgical Population: A Randomized Controlled Trial. JAMA: The Journal of the American Medical Association 2004; 291: 79-87

16.    Maragakis LL, Cosgrove SE, Martinez EA, Tucker MG, Cohen DB, Perl TM: Intraoperative Fraction of Inspired Oxygen Is a Modifiable Risk Factor for Surgical Site Infection after Spinal Surgery. Anesthesiology 2009; 110:

17.    Myles PS, Leslie K, Chan MTV, Forbes A, Paech MJ, Peyton P, Silbert BS, Pascoe E, the ENIGMA Trial Group: Avoidance of Nitrous Oxide for Patients Undergoing Major Surgery: A Randomized Controlled Trial. Anesthesiology 2007; 107:

18.    Meyhoff CS, Wetterslev J+, Jorgensen LN, Henneberg SW, H+©gdall C, Lundvall L, Svendsen PE, Mollerup H, Lunn TH, Simonsen I, Martinsen KR, Pulawska T, Bundgaard L, Bugge L, Hansen EG, Riber C, Gocht-Jensen P, Walker LR, Bendtsen A, Johansson G, Skovgaard N, Helt+© K, Poukinski A, Korshin A, Walli A, Bulut M, Carlsson PS, Rodt SA, Lundbech LB, Rask H, Buch N, Perdawid SK, Reza J, Jensen KV, Carlsen CG, Jensen FS, Rasmussen LS: Effect of High Perioperative Oxygen Fraction on Surgical Site Infection and Pulmonary Complications After Abdominal Surgery. JAMA: The Journal of the American Medical Association 2009; 302: 1543-50

19.    Qadan M, Akca O, Mahid SS, Hornung CA, Polk HC, Jr.: Perioperative Supplemental Oxygen Therapy and Surgical Site Infection: A Meta-analysis of Randomized Controlled Trials. Archives of Surgery 2009; 144: 359-66

20.    Al-Niaimi A, Safdar N: Supplemental perioperative oxygen for reducing surgical site infection: a meta-analysis. Journal of Evaluation in Clinical Practice 2009; 15: 360-5

This review copyright Patrick Neligan 2012. All rights reserved. Do not reproduce without permission.

Acute Respiratory Distress in the Recovery Room (tutorial)

Clinical Scenario: A 57 year old male undergoes upper abdominal surgery. He refused an epidural. The intraoperative course was uneventful. He was given 2mg hydromorphone in the OR. He was extubated, breathing 360 ml tidal volumes; arousable. Shortly after arrival to the recovery room, the patient develops acute respiratory distress. His respiratory rate increases to 33 breaths per minute, SpO2 is 92%, heart rate increases to 110 beat/min, blood pressure 98/50 mmHg. On examination, his pupils are pinpoint but reacting, he is moving air into both of his lungs but there is little air entry into his lung bases.

A non-rebreather facemask is placed: his SpO2 remains 92%.

1. Identify the problem

What is the principle diagnosis?

The patient is clearly in acute respiratory distress; however the cause and reversibility of the problem are unclear. It is imperative that the bedside clinician have a systematic approach to diagnosis and management. The cause of the problem may lie at any stage in the process of initiating a breath to exchanging gas. This tutorial focuses on the diagnosis.

Patients in recovery room with acute respiratory distress have one or more of the following three problems: failure to ventilate, as characterized by a high PaCO2, failure to oxygenate, as characterized by low PaO2, or failure to maintain their airway (figure 1). All three may co-exist: for example, a patient that receives excess opioids my hypoventilate, obstruct their airway due to opioids and carbon dioxide narcosis, and become hypoxic due to absorption atelectasis, failure to replenish alveolar oxygen and alveolar CO2 buildup. Nevertheless, the primary problem is failure to ventilate, due to central loss of respiratory drive. Hence, to make a diagnosis, one needs to identify the primary problem.

 Figure 1: Mechanisms of Acute Respiratory Distress in recovery room

There are three major components to the respiratory apparatus:

  1. Central chemoreceptors: in the brainstem that detect carbon dioxide and initiate the respiratory pump. This requires an intact brainstem and cervical nerve roots.
  2. The respiratory pump: the phrenic nerves (and on occasion the intercostals nerves) initiate diaphragmatic contraction. This requires and intact neuromuscular junction and sufficient diaphragmatic muscular tissue to increase the volume of the thoracic cavity. This leads to increased negative pressure within the pleura, stretching the alveoli. Flow of gas into the alveoli is known as ventilation.
  3. Alveolar-capillary interface: gas must flow across the alveolar capillary interface to enter and leave the blood. This is known as ventilation perfusion matching and is reliant on alveolar gas volume (particularly in end expiration – the functional residual capacity) and pulmonary blood flow.

The problem is either central – a problem of respiratory drive, peripheral – a problem of the respiratory pump, large airway – a problem of gas transfer, or alveolar – a problem of gas exchange (figure 2).

  1. Central Ventilation: the neurologic system is not activating respiration in response to an increase in arterial CO2 tension
  2. Peripheral Ventilation: the thoracic pump (chest and diaphragm) is not effective in guaranteeing adequate minute ventilation.
  3. Gas Transfer: air does not pass effectively from the upper to the lower airway due for example to increased airway resistance.
  4. Gas Exchange:
    1. Gas does not to pass effectively from alveoli to capillaries due to a pathologic process in the interstitial space (diffusion defect).
    2. Ventilation is being wasted – alveoli are being ventilated but not perfused: dead space ventilation or more air than the blood can utilize (high ventilation/perfusion (V/Q) ratio the extreme version being dead space ventilation).
    3. Blood flow is inadequately utilized and blood is passing through the lungs without coming into contact with aerated alveoli: perfused but not ventilated – shunt or ventilation falls behind blood flow (low V/Q ratio the extreme version being right to left shunt).

2. Understand the problem

What is the mechanism of injury?

Ventilation Failure

Failure to ventilate is the most common cause of acute respiratory distress in the recovery room. It is characterized by reduced alveolar ventilation which manifests as an increase in the PaCO2 > 50 mmHg (6.5kPa). The best method of classifying this is to follow the respiratory  pathways from the brainstem to the alveoli, and then ask whether a pathology exists at each particular site. Often patients have multiple problems: e.g. narcosis, pulmonary edema, pleural effusion, obesity


Central: loss of ventilatory drive due to general anesthetic agents (propofol principally), benzodiazepines, narcosis, stroke or brain injury

Spinal: spinal or epidural anesthesia; spinal cord injury, cervical – loss of diaphragmatic function, thoracic – loss of intercostals.

Peripheral: phrenic nerve injury in neck or thoracic surgery


Persistent neuromuscular blockade; diaphragmatic trauma; myopathic disorders – myasthenia gravis (patient post op thymectomy).

Anatomical Problems

Chest Wall – flail chest; intra-abdominal hypertension (abdominal packs placed).

Pleura – pleural effusions, pneumothorax (patient post op thoracic or retroperitoneal surgery: nephrectomy, abdominal aortic aneurysm, esophagectomy).

Airways – airway obstruction: laryngeal edema, inhalation of a foreign object (tooth or throat pack), bronchospasm.

Oxygenation Failure

Oxygenation failure occurs at a microscopic level at pulmonary capillary-alveolar interface. Two different injuries can occur at this level, either individually or in combination:

Diffusion abnormality – thickening of the alveoli (pulmonary fibrosis). There is an obstruction to effective gas exchange due to material in the interstitial space. The patient will have an antecedent history of hypoxemia.

Ventilation/Perfusion Mismatch: Dead Space Ventilation (or high V/Q): alveoli are ventilated but not perfused. This is unusual in the extubated patient, an usually results from significant hypovolemia


Figure 2 Causes of Respiratory Failure

Ventilation Perfusion Mismatch (figure 3): this occurs when lung units well perfused but poorly ventilated. The extreme version is right (as in right side of the heart) to left shunt (blood flows through the lungs without coming into contact with aerated lung tissue. This lung injury is resistant to oxygen therapy. This frequently occurs in patients that have upper abdominal or chest surgery secondary to segmental lung collapse – atelectasis. Atelectasis may actually be worsened by oxygen therapy, due to rapid reabsorption.

Less severe, and usually oxygen sensitive, ventilation-perfusion mismatch is the inevitable consequence of major surgery. The time constants in many lung units are altered due to edema in the lung tissue and secretions in the major and minor airways. This results in an alteration of the dynamics of gas transport: alveolar oxygen tension is slower to be replenished, carbon dioxide is more slowly removed.

Figure 3: Ventilation-Perfusion Mismatch. Alveolar unit A has normal ventilation and perfusion, hence the pulmonary capillary (arterial side) oxygen tension (PcO2) is 100mmHg (13kPa). Unit C is ventilated but not perfused. It does not contribute to gas exchange. Unit B is partially ventilated, but due to it’s long time constant (due to secretions), the alveolar oxygen tension is below normal, and the PcO2 is reduced to 70mmHg (9kPa). When all lung units are accounted for, the result is hypoxemia (PaO2 70mmHg/9kPa)

Figure 4: Oxygen therapy effectively treats ventilation perfusion mismatch by increasing the fraction of gas in the alveolus that is oxygen, thus increasing the PAO2 (alveolar O2). It also reverses pulmonary vasoconstriction and reduces dead space.

Figure 5: Right to left intrapulmonary shunt: in this example, 50% of the pulmonary circulation is flowing thru collapsed lung tissue. Because hemoglobin can only be saturated to 100%, regardless of the quantity of oxygen that is delivered to normal lung tissue, it is not possible to compensate for the intra-pulmonary shunt.

3. Differential diagnosis / Work the problem

How do you make the diagnosis?

Acute respiratory failure is usually a problem of either failure to oxygenate, as characterized by a low PaO2, or failure to ventilate, as characterized by a high PaCO2. Where hypoxemia and hypercarbia co-exist, oxygenation should be considered the primary problem.


Figure 6: Assessing the Patient with Acute Respiratory Failure

The key to making the diagnosis is to look at the patient’s breathing pattern. If the patient is taking slow shallow breaths, with normal synchrony between opening of the mouth to inhale and movement of the chest outwards and downwards, then the problem is most likely ventilatory failure secondary to central respiratory depression. This most commonly results from opioid administration, but may also follow the administration of midazolam/lorazepam or discontinuation of a propofol infusion.

If the patient is taking rapid shallow breaths, the problem is either ventilatory failure secondary to a peripheral problem or oxygenation failure secondary to ventilation-perfusion mismatch. The key to separating the two is the clinical circumstance and the presence or absence of hypoxemia (low SpO2 or requirement for high FiO2). In the absence of hypoxemia, a neuromuscular problem should be considered – such as residual neuromuscular blockade or a dense epidural block that paralyses the intercostal muscles. One also sees this pattern in patients with low physiologic reserve, the malnourished and the critically ill. In patients that have undergone thoracic surgery or retroperitoneal surgery, a high clinical suspicion for pneumothorax should be considered. This is characterized by hypoxemia, unilateral breaths sounds, and, in severe cases, hypotension.

Rapid shallow breathing with hypoxemia is caused by ventilation perfusion mismatch. This is usually caused by retained secretions and/or atelectasis. This most commonly occurs in patients that have undergone abdominal surgery, are morbidly obese or have been positioned intraoperatively in the Trendelenberg position.

Pulmonary embolism should be suspected in patients that have undergone pelvic or hip surgery with rapid shallow breathing and hypoxemia, associated with tachycardia and hypotension.

An obstructed breathing pattern is suggestive of upper or middle airway pathology. The problem is caused by central loss of pharyngeal tone, and soft tissue obstruction (associated with depressed level of consciousness and anesthesia) or mechanical obstruction to the airway, above, at the level of or below the glottis. Classically the patient has nasal flaring, supraclavicular or intercostal retraction, and a see-saw chest movement: the chest moves inwards as the diaphragm descends. The patient may have inspiratory stridor (supraglottic obstruction), expiratory stridor (glottic or subglottic obstruction) or expiratory wheeze (bronchospasm). Typically hypoxemia is a late complication of airway obstruction. This is important as hypoxia may be rapidly followed by bradycardia and asystole.

4. Solve or resolve the problem

This patient has many risk factors for acute respiratory distress. Does he have ventilatory failure? Quite possibly – he may be narcosed from excessive interoperative opioids. He may be hypoventilating due to splinting (upper abdominal pain due to surgical incision) or persistent partial neuromuscular blockade. He may have upper airway obstruction, due to loss of pharyngeal tone, obstruction with a bite block, laryngeal edema or laryngospasm. He may have severe bronchospasm, and inhaled foreign object (such as a tooth) obstructing a major bronchus, or the presence of blood or gastric contents aspirated from the upper airway. He may have lower airway collapse due to hypoventilation and or absorption atelectasis, diffusion hypoxia (due to oxygen being displaced by nitrous oxide in the alveoli) or alveolar fluid, due to excessive intravenous administration.

Working the problem:

Step 1: Is this failure to oxygenate or failure to ventilate?

The patient has rapid shallow breathing with hypoxemia – this is failure to oxygenate, it may be secondary to peripheral ventilatory failure, or primary to V/Q mismatch.

Step 2: Is this peripheral ventilatory failure or primary V/Q mismatch?

The patient has bilateral air entry into the upper segments of the lungs, with little air entry into the bases. This is primary V/Q mismatch secondary to atelectasis. The patient has an intra-pulmonary shunt, evidenced by the lack of responsiveness to oxygen therapy.

Step 3: How is the diagnosis confirmed?

The diagnosis may be accepted, clinically (there is sufficient clinical suspicion in this case) or confirmed by chest x-ray and arterial blood gas sampling.

Step 4: What is the initial management of this patient?

The patient should nursed in the upright or seated position – the effect of gravity is to recruit lung tissue and increase functional residual capacity. The patient should be encouraged to cough, to mobilize secretions. Consideration should be given to devices that assist in lung recruitment such as incentive spirometry or the use of non-invasive positive pressure ventilation. If the problem worsens or fails to resolve, the patient should be re-intubated and lung recruitment achieved using an ICU grade mechanical ventilator.

5. Conclusions

  1. The assessment of the patient with acute respiratory distress involves taking a history, examining the patient and quantifying the degree of respiratory injury.
  2. This involves determining whether the problem is failure to ventilate, failure to oxygenate or failure to maintain the airway.
  3. Failure to maintain the airway leads to failure of gas flow and ultimately hypoxemia and hypercarbia. The problem is either central loss of airway patency or mechanical airway obstruction.
  4. Failure to oxygenate is caused by ventilation perfusion mismatch: the patient typically has a rapid shallow breathing pattern.
  5. Failure to ventilate is caused by a problem in the central nervous system or a problem with the thoracic pump: the patient typically has a slow shallow breathing pattern.
  6. Failure to ventilate is an ominous sign.
  7. Look for an immediately reversible cause of failure to ventilate – such as narcosis, deep sedation or persistent neuromuscular blockade.
  8. In the absence of a reversible cause, positive pressure ventilation is required.


Figure 7: Failure to Oxygenate vs Failure to Ventilate

This article is entirely the work of Patrick J Neligan MA MB FCAI FJFICM. No part of this article or its illustrations may be reproduced without the author’s permission. Select illustrations were developed in conjunction with Maurizio Cereda MD. © PJN 2012