More dogma, or is it actually useful? Any time a chest tube (tube thoracostomy) is inserted, we automatically order a chest x-ray. Even the ATLS course recommends obtaining an image after placement. But anything we do “automatically” is grounds for critical analysis to see if there is a valid reason for doing it.
A South African group looked at the utility of this practice retrospectively in 1004 of their patients. They place 1042 tubes. Here are the factoids:
- Patients were included if they had at least one chest x-ray obtained after insertion
- Patients were grouped as follows: Group A (10%) had the tube inserted on clinical grounds with no pre-insertion x-ray (e.g. tension pneumothorax). Group B (19%) had a chest x-ray before and had ongoing clinical concerns after insertion. Group C (71%) had a chest-xray before and no ongoing concerns.
- 75% of injuries were penetrating (75% stab, 25% GSW), 25% were blunt
- Group A (insertion with pre-x-ray): 9% had post-insertion findings that prompted a management change (kinked, not inserted far enough)
- Group B (ongoing clinical concerns): 58% required a management change based on the post-x-ray. 33% were subcutaneous or not inserted far enough (!!)
- Group C (no ongoing clinical concerns): 32 of 710 (5%) required a management change, usually because the tube was too deep
The authors concluded that if there are no clinical concerns (tube functioning, no clinical symptoms) after insertion, then a chest x-ray is not necessary.
Bottom line: But I disagree with the authors! Even with no obvious clinical concerns, the tube may not be functioning for a variety of reasons. Hopefully, this fact would then be discovered the next day when another x-ray is obtained. But this delays the usual progression toward removing the tube promptly by at least one day. It increases hospital stay, as well as the likelihood of infection or other hospital-associated complication. A chest x-ray is cheap compared to a day in the hospital, which would potentially happen in 5% of these patients. I recommend that we continue to obtain a simple one-view chest x-ray after tube insertion.
One of the cornerstones of allopathic medicine is the use of drugs to treat disease conditions. And unfortunately, one of the side effects of using drugs to treat problems is the production of side effects(!).
In trauma care, even something as simple as treating pain from an injury can create major problems. Give a narcotic pain medication. The patient gets nauseated and vomits. Try a different narcotic. The patient develops constipation. Give stool softeners and cathartics. Diarrhea. Then pseudo-obstruction develops. Give neostigmine. The patient becomes bradycardic. Give… well, you get the picture.
How common are side effects? Very! Did anyone see the first TV commercials for Chantix, the smoking cessation drug? It was about 3 minutes long because of the long list of side effects that were described. I’m surprised anyone was willing to risk them just to stop smoking cigarettes.
A recent study looked at the number of side effects listed on the labels of 5,602 medications approved by the FDA. There were a grand total of 534,125 adverse drug effects described in the packaging. Some interesting statistics:
- The number of adverse effects for ranged from 0 to 525(!) for a single drug
- The median number of adverse effects was 49, the average was 70
- Drugs with the most side effects are used in neurology, psychiatry and rheumatology
- Newer drugs had significantly more adverse effects than older ones
It’s certainly easy to bash pharmaceutical companies on their products. But some of these findings may be due to more rigorous testing and monitoring, as well as nuances in the populations in which these drugs are used.
Bottom line: Drugs are chemicals! Each chemical has a number of effects, some of which are desirable, and some of which are not. The drug companies choose to market a drug based on one desired effect (e.g. control of nausea). Just remember, when you give that medication, you will probably get the desired effect, but you will just as likely also get some of the other 69 possible side effects. Be prepared, and prescribe sensibly.
Reference: A quantitative analysis of adverse events and “overwarning” in drug labeling. Arch Int Med 171(10):944-946, 2011.
Open fracture dogma has mandated management of these injuries within an 8 hour window. Over the past several years, there has been a growing number of good papers that dispute this fact. As is the norm, many are retrospective in nature, or meta-analyses of retrospective papers.
Recently, a paper was published that detailed a (small) prospective and multi-center study (3 hospitals in Canada) looking at deep infection, Gustillo grade, antibiotics, and time to treatment. My hopes were raised!
Here are the factoids:
- 939 patients were screened, but only 736 were actually enrolled
- Only 482 completed the entire study (>90 days clinical followup and an interview after 1 year). Others with less clinical followup were still included and analyzed.
- Information on fracture grade, time to antibiotic administration, time to OR, and development of deep infections were recorded. Cellulitis and pin site infections were not considered.
- Time to antibiotic administration ranged from 1 hour to 10 hours (!!?)
- Time to OR ranged from 6 to 13 hours
- 46 patients developed deep infections, and 56 had cellulitis or pin site infections
- Of those who developed infections, there was no clear association with time to OR
- Also in those with infections, antibiotics were given after about 3 hours, vs 2.5 hours in those without.
The authors concluded that neither time to antibiotic administration nor time to surgery made any difference on deep tissue infections. But should I believe them?
Bottom line: SLOPPY! If you just read the abstract you might believe the wrong thing. This paper cobbled together surgeons at 3 different centers and let them do their own thing. The researchers just observed the management that these fellowship trained surgeons chose. No guidelines. No protocols. The variability of practice in this study leaves me flabberghasted. The median time to antibiotic administration was 3 hours, with some waiting up to 10 hours! The median time to OR was 9 hours, not so far off the 8 hour mark the everyone seems to look at. No wonder they couldn’t find any differences.
Give antibiotics early. Get to the OR in a reasonable amount of time, preferably using the Gustillo grade to take high grades there sooner. And keep your eye on the literature for papers that are much, much better than this one!
Reference: Time to Initial Operative Treatment Following Open Fracture Does Not Impact Development of Deep Infection: A Prospective Cohort Study of 736 Subjects. J Orthop Trauma 28(11):613-619, 2014.
EMS policy and the trauma center verification process requires that all trauma patients delivered to a trauma center must have a copy of the EMS run sheet. Two parameters that are commonly used to monitor performance improvement (PI) in EMS are:
- accurate record of scene physiology (SBP, HR, RR, GCS)
- request by on-scene BLS for ALS assistance
A study looked at the impact of those criteria on patient survival. A total of 4744 patients from the National Trauma Data Bank were analyzed.
Physiologic data: About 28% had at least one missing physiologic data point, with respiratory rate being most commonly missed. They found that the mortality in the group with missing data was over twice as high (10.3%) as it was in the group with complete date (4.5%).
BLS call for ALS assistance: This assist was called for in 17% of cases. These cases were less likely to involve penetrating injuries and more likely to involve car or motorcycle crashes. Injury Severity Score was the same. Eventual patient mortality was the same for BLS calling ALS and ALS response alone.
So why does failure to record physiologic data translate into higher mortality? The initial response may be that the patient was sicker, and so they needed more intense care during transport with less time to record vitals. However, the researchers controlled for this and found it was not a factor. Other issues that may be a factor are EMS training and proficiency, leadership at the scene and enroute, and available staff and resources, among other things.
The researchers speculate that documentation might be a good global measure of appropriate or inappropriate prehospital care that rolls all of these possible factors into one easily identifiable audit filter. They recommend that this be used to focus performance improvement efforts and hopefully improve survival.
Bottom line: I have visited a number of states where EMS often does not leave their run sheet at all! I recommend that the results of this study be taken to heart and used to help persuade EMS programs to get religious about recording complete vital signs and leaving the run sheet at the trauma center every time a patient is delivered. Documentation should be evaluated regularly, and all cases with any missing vital signs should be reviewed closely. Trauma Center PI programs should work with EMS to analyze this data and look for the patterns that increase mortality.
Reference: Lack of Emergency Medical Services documentation is associated with poor patient outcomes: a validation of audit filters for prehospital trauma care. Journal of the American College of Surgeons, 210(2):220-227, 2010.
This paper was presented at EAST in 2013, and this is an update of that work using the entire manuscript which has now been published.
The standard of care in most high level trauma centers is to involve neurosurgeons in the care of patients with significant traumatic brain injury (TBI). However, not all hospitals that take care of trauma patients have immediate availability of this resource. The University of Arizona at Tucson looked at management of these patients by their acute care surgeons.
The authors did a retrospective cohort study of patients at their center who had a mild TBI and positive head CT, managed with or without neurosurgery consultation, over a two year period. They matched the patients with and without neurosurgical consultation for age, GCS, AIS-Head and presence of skull fracture and intracranial hemorrhage.
A total of 90 patients with and 180 patients without neurosurgical involvement were reviewed. Here are the factoids:
- Hospital admission rate was identical for both groups (87-89%)
- ICU admission was significantly higher if neurosurgeons were involved (20% vs 44%)
- Repeat head CT was ordered more than 3 times as often by neurosurgeons (20% vs 86%)
- Post-discharge head CT was ordered more often by neurosurgeons, but was not significantly different (5% vs 12%)
- There were no surgical interventions, in-hospital mortality, or readmissions within 30 days in either group.
- Cost of the hospital stay was significantly increased if neurosurgery was consulted.
Bottom line: Can surgeons safely manage select patients with intracranial injury? Granted, this is a small, retrospective study, but the answer is probably yes. The majority of patients with mild to moderate TBI with small intracranial bleeds or skull fractures do well despite everything we throw at them. And it appears that surgeons use fewer resources managing them than neurosurgeons do. The keys to being able to use this type of system are to identify at-risk patients who really do need a neurosurgeon early, and having a quick way to get the neurosurgeon involved (by consultation or hospital transfer). Having a specific practice guideline for management is essential as well. As neurosurgery involvement in acute trauma declines, this concept will become more and more pertinent.
Reference: The acute care surgery model: managing traumatic brain injury without an inpatient neurosurgical consultation. J Trauma 75(1):102-105, 2013.