Date: March 20th, 2018
Reference: Sawa et al. Risk factors for adverse outcomes in older adults with blunt chest trauma: A systematic review. CJEM March 2018
Guest Skeptic: Dr. Christina Shenvi is an Emergency Physician at University of North Carolina. She is fellowship trained in Geriatric Emergency Medicine and has a podcast called GEMCAST. Christina loves serving as the assistant residency director, writing things, reading things, teaching people, and having kids.
Case: An 85-year-old woman with a history of chronic obstructive pulmonary disease (COPD), congestive heart failure (CHF), and type-2 diabetes (DM-2), and her 65-year-old otherwise healthy daughter present to the emergency department after a car accident. It was a low speed motor vehicle collision (MVC) in which they rear-ended a stationary car. However, they were both unrestrained. They both have a Glasgow Coma Scale (GCS) of 15 on arrival and are complaining of chest pain. The physician in the emergency department is deciding what imaging to obtain and will ultimately have to decide the disposition.
Background: Older adults who sustain a blunt traumatic injury to the chest are at higher risk for greater injury severity for a given mechanism, longer lengths of stay, loss of independence, and higher morbidity and mortality than younger patients.
However, it is not a homogeneous population. Some older adults will do relatively well, while others will be at particularly high risk for adverse outcomes.
Clinical Question: What are the main factors that are associated with adverse outcomes in older adults with blunt chest trauma?
Reference: Sawa et al. Risk factors for adverse outcomes in older adults with blunt chest trauma: A systematic review. CJEM March 2018
- Population: Older adults defined as 65-years of age and over. Blunt chest trauma defined as “blunt chest injury resulting in chest wall contusion or rib fractures, with or without immediate life-threatening injury to the lungs or other organ systems.”
- Review articles, case reports, or case-series
- Population not meeting inclusion criteria
- Did not analyze risk factors for outcome of interest
- Did not examine an outcome of interest
- Intervention: Not applicable
- Comparison: The comparison was prognostic factors for outcomes of interest. The factors were grouped into three categories:
- Patient Factors: Any underlying features, conditions, or demographics present before the injury such as age, sex, co-morbidities.
- Disease Factors: Any risk factors related to the traumatic event such as number of rib fractures, flail chest, mechanism, other associated trauma.
- Institutional Factors: These had to do with the hospital and emergency department such as: was the patient seen by a trauma team, was there a multi-disciplinary team, and were there any adverse events that occurred.
- Primary Outcomes:
- Morbidity: Pneumonia, intubation and intensive care admission
- Mortality: In-hospital or 30-day
- Emergency Department Recidivism
- Length of Stay (LOS)
- Quality of Life (QOL)
- Loss of Independence: Defined as discharge to a higher level of community care than they were admitted from.
Authors’ Conclusions: “While blunt chest wall trauma in older adults is relatively common, the literature on prognostic factors for adverse outcomes in this patient population remains inadequate due to a paucity of high quality studies and lack of consistent reporting standards.”
Quality Checklist for Systematic Review of Prognostic Studies:
- The prognostic question is clinically relevant for emergency department patients? Yes.
- The individual study patients were sufficiently homogeneous with respect to prognostic risk for the outcome? No
- The individual study assessment for the outcome used objective, reproducible, and unbiased criteria? Unsure
- The individual study period of follow-up was sufficiently long and complete? Yes and No.
- The search for studies was detailed and exhaustive? No
- The methodological quality of primary studies were assessed for common forms of prognostic research bias? Yes
- The assessment of studies were reproducible? Unsure
- There was low heterogeneity for estimates of sensitivity or specificity? No
- The summary prognostic accuracy is sufficiently precise to improve upon existing clinical decision-making models? No
Key Results: Their search resulted in 13 studies that met all their criteria, representing a total of over 79,000 patients. There were eleven studies from the US, one from Egypt, and one from Israel.
All the studies were retrospective cohort studies, primarily relying on trauma registries. Of note, one of the studies accounted for the vast majority of patients and included 67,659 patients. At the other end of the spectrum, the study from Egypt had 39 patients, and the study from Israel had 77 patients.
There was too much heterogeneity and the authors were not able to perform a meta-analysis. However, the overall mortality in the studies was 8.4% and the complication rate was 26.5%. They also wanted to look at emergency department recidivism and quality of life, but none of the studies included this information.
Patients are at higher risk of adverse outcomes with increasing age, increasing rib fractures, greater burden of underlying chronic disease, and if they required early intubation or had worse oxygenation on admission.
- Patient Factors: One study looked at Age >= 80, which had a statistically significant OR of 2.37, another study found an OR of 1148.5 (CI 184.9-7132.6) with “age” but the paper didn’t clarify what it meant by “age”. The largest study in the review had an OR of mortality of 1.059 for death. So, for example, if your OR of death were 1 for a 65yo, that would give you an OR of 1.77 at age 75 and 2.36 at 80. Those studies have a pretty big difference in the OR they report. The study that had the OR over 1000 was a study from the US of 1621 patients and had low risk of bias. One study showed OR 5.7 in patients with congestive heart failure
- Disease/Injury Factors: Several studies looked at number of rib fractures, and there was a statistically significant increase in mortality, but there was too much heterogeneity to do any sort of meta-analysis. Certainly, with lots of rib fractures (over 8), one study had an OR of 1.54 for mortality
- Institutional Factors: Only one study looked at whether admission to a level 1 trauma center had an effect on mortality, and it did, with an OR 4.5. This adjusted for age, ISS, need for intubation, and pre-existing CHF.
- Patient Factors: Outcomes of myocardial infarction, pneumonia, pulmonary contusion and effusion were increased in patients with pre-existing cardiopulmonary disease (OR 8.2). General “pulmonary complications”, pneumonia, and contusions, were increased with diabetes (OR 5.7 for the first and 11.5 for the latter 2). Other significant risk factors: COPD, protein calorie malnutrition, and use of ambulatory assist devices. Per year increase in age, the need for mechanical ventilation had an OR of 1.004, and for pneumonia 1.007
- Disease/Injury Factors: More pneumonia, pulmonary embolism, or other pulmonary complications, with lower oxygen saturation on admission, more rib fractures, earlier need for mechanical ventilation, and with each increase in rib fractures, or each point increase in ISS.
- Institutional Factors: Pleural effusion or pneumonia increased with use of epidural vs. IV analgesia (OR 3.3) after adjusting for ISS and pre-existing disease.
- Length of Stay:
- Patient Factors:Pre-existing cardiopulmonary disease
- Disease/Injury Factors:Need for mechanical ventilation, number of rib fractures, or over five rib fractures.
- Institutional Factors: Use of IV vs. epidural analgesia (5.6 vs 8.6 days). Protocolized care of rib fractures vs pre-protocol care reduces LOS (7.1 vs 8.2 days).
- Discharge to an Institution: Only one study looked at this and found those with lower mean vital capacities within 48 hours of admission were more likely to be discharged to an extended care facility.
This was a pretty solid systematic review given the limitations of the available evidence. They adhered to the PRISMA and MOOSE guidelines for systematic reviews and meta-analyses.
They described various kappas, confidence intervals, contingency tables, risk of bias, non-weighted average of reported means, and other statistical terms to describe their work.
1) Search Strategy: They used an expert librarian, searched two electronic databases and hand searched the bibliographies of all the included studies. However, they did not do a great job of searching the grey literature. They did not mention going through conference abstracts or reaching out to experts in the field (like Chris Carpenter). They also restricted inclusion to abstracts in English and French.
2) Heterogeneity: The studies, unfortunately had a fair amount of heterogeneity. Some measured risk of outcomes based on age or rib fractures in a binary fashion (ex: age >64 or age >80, or <8 or >8 rib fractures) while others measured the effect per year or per fracture. There was also heterogeneity in the trauma burden. The mean Injury Severity Score (ISS) score of studies ranged from 6.9 to 19.4, and four of the studies did not even report ISS, and several of them did not adjust their predictive factors for the patient ISS. Studies also lumped their outcomes, such as including “pneumonia, pleural effusion, and intubation” as a single outcome. This is not very helpful or patient-centered. Having a little pleural effusion is very different from having to be intubated.
3) Follow-up: Some of the studies looked at 30-day mortality and LOS, which are pretty standard outcome measures. They did want to measure QOL and loss of independence, but the included studies did not measure these outcomes. Both are very important patient-oriented outcomes, and would need to be measured further out, but just haven’t been the focus of studies so far and require more intensive follow-up. As with all systematic reviews, the authors were limited by the available published evidence.
4) Study Bias: They looked at risk of bias in the study participation, attrition, measurement of prognostic factors and outcomes, study confounding, and statistical analysis and presentation. Unfortunately, most of the studies did have high risk of study confounding and risk of bias in statistical analysis. However, the study that had the largest number of patients, (67,559/79,313 or 85%) had low risk of bias in all categories.
5) Associations and Strength of Associations: All the studies included in the systematic review were retrospective cohort studies. This means we are only talking about associations between the reported risk factors and outcomes of interest. In addition, some of the studies did not perform a multivariate analysis. This limits the ability to interpret the relative strength of associations across studies.
Comment on Authors’ Conclusion Compared to SGEM Conclusion: We essentially agree with their conclusions.
SGEM Bottom Line: Age, Injury Severity Score, number of rib fractures, and chronic disease are associated with higher morbidity, mortality, and length of stay in blunt trauma patients 65 years or older.
Case Resolution: The astute emergency physician obtains a chest x-ray in both patients. They are both read as negative. However, the 85-year old patient has marked tenderness over the right lower chest wall. A CT is obtained that shows three rib fractures. The patient is admitted given her very high risk for pneumonia or other respiratory complications. She receives pain medication and pulmonary therapy and is discharged four days later.
The 65-year old patient has some bruising over the chest wall but no significant point tenderness, no splinting, no respiratory distress, and no hypoxia. Since she is generally healthy, ambulatory, oxygen saturation is good, and is unlikely to require admission even if she has an isolated rib fracture, she is discharged with acetaminophen and does well recovering at home.
Clinical Application: It is challenging to know how to use this information prospectively, as for example, their inclusion requires the diagnosis of chest trauma, such as rib fractures. That implies that imaging has already been performed. But it’s not clear, for example whether these rib fractures were diagnosed on X-ray, which misses a lot of fractures, or if it was based on CT. So, it’s not necessarily as helpful as one would hope in making the decision about what imaging modality to choose, or whether imaging is needed.
Another problem is that none of the ten-potential patient and disease factors for morbidity are likely to be modifiable. Certainly, a patient’s age is not modifiable and hence the theme music…holding back the years…it is not possible.
However, it is reasonable to apply this data to clinical care, primarily by keeping an awareness that patients who are older, have more rib fractures, and more co-morbidities who present with blunt trauma are at higher risk for morbidity and mortality. This may be helpful to guide getting more imaging or helping guide disposition. However, there isn’t a really clear cutoff for things, such as how many rib fractures is ok, or how many rib fractures plus co-morbidities are ok?
At the end of the day, just be more cautious in your care of older adults with blunt trauma. Morbidity and mortality is associated with increase with age, ISS, number of rib fractures, and the presence of underlying chronic disease.
What Do I Tell the Patients? I am concerned that since you have multiple rib fractures, and you have underlying lung disease, that you are at high risk for developing pneumonia, or having worsening difficulty breathing. I think you should be admitted so that we can continue to make sure you are getting enough oxygen, and that your pain is well controlled, so that we can closely monitor and treat you if you do develop difficulty breathing or a lung infection.
For the 65-year woman I would say we have performed an X-ray, we do not see any rib fractures and since your pain seems to be reasonable controlled with acetaminophen, I would recommend to take this medication as needed at home. However, if you start to have worse pain, cough, fever, or trouble breathing, I want to you come back to the emergency department to be re-evaluated.
Keener Kontest: Last weeks’ winner was…No one knew furosemide does not have a role in the initial treatment of pulmonary edema. It has a delayed action that is designed to work on an organ system that is not being perfused when the patient is under stress. When the flight or fight reflex kicks in the blood flow to the kidney’s is shut down. Nitroglycerine and Captopril provide acute preload and afterload reduction which both help the heart to unload the fluid filled lungs. And the sauna thing, there are two studies in CHF patients with Swann-Ganz catheters in place shows a drop in PCWP when the patients were vasodilated in a hot sauna room.
Listen to the SGEM podcast on iTunes to hear this weeks’ question. If you know the answer send an email to TheSGEM@gmail.com with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.
“The Philosopher” by J.G.Woodward, from The Scientific Monthly, March 1945.
- Say on, philosopher, what have you found?
- What weighty, abstruse tome would you expound…
- When you say, “Yes,” your colleague answers, “No.”
- And thus the ponderous volumes swell and grow…
- And Truth is something Man may never know.
- Yet seek! Perhaps not even this is so.