• 
• 

Three Part Question

In [persons who are at risk for cold injuries] does [application of emollients] prevent [frostbites] compared to [no emollient application]?

Clinical Scenario

A 23 year old man went on a skitrip in an extreme cold and windy environment. He protected his face against cold injuries by applying of a protecting emollient, but he got frostbite in the face anyway. When arriving at the emergency department, he asks you whether the use of emollients isn’t a good protection against frostbites.

Search Strategy

The Cochrane Library using the Wiley interface from date of inception to 10th May 2012: MeSH descriptor Frostbite explode all trees. 11 hits
MEDLINE from date of inception to 10th May 2012 using the PubMed interface: ("Frostbite"[Mesh] OR "Cold Temperature"[Mesh]) AND ("Ointments"[Mesh] OR "Emollients"[Mesh]). 36 hits
Embase and MEDLINE from date of inception to 10th May 2012 via the Embase.com interface: ('cold injury'/exp OR 'cold'/exp OR frostbite) AND ('emollient agent'/exp OR 'ointment'/exp). 53 hits

Ex vivo experiments and animal studies were excluded.

Search Outcome

100 papers were found of which one describes an ex vivo experiment, one compares different ointments and 95 articles were irrelevant. The remaining three papers are shown in the table.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Lehmuskalio 2000 Finland	24 voluntary and healthy male test persons (mean age 26 years, range 19 - 48 years) with a non-medicated emollient on one half of the face at 50 g/m², while the other half acted as an untreated control (Ambient temperature -15°C, wind 3 m/s against the face)	Intervention study (within subjects design) (LOE 2b)	Percentage of subjects with a higher facial skin temperature at 5, 10, 15, 20 and 25 min when using emollients vs no emollients (measurement with thermistors)	28%, 95% CI: 18-38	Small study; not randomized; problems with objective measurement of the skin temperature when ointments are applied (however results are mostly consistent when measured with an IR method); 4 different non-medicated ointments were tested and evaluated as sum of emollients. However, the inter-emollient difference was not significant (x2 = 5.85; degree of freedom = 8;p > 0.50).
			Percentage of subjects with a lower facial skin temperature at 5, 10, 15, 20 and 25 min when using emollients vs no emollients (measurement with thermistors)	51%, 95% CI: 40-62
			Percentage of subjects with a higher facial skin temperature at 5, 10, 15, 20 and 25 min when using emollients vs no emollients (measurement with infrared scanner; values ≥ 0.7°C were considered as thermal difference)	12%, 95% CI: 5-19
			Percentage of subjects with a lower facial skin temperature at 5, 10, 15, 20 and 25 min when using emollients vs no emollients (measurement with infrared scanner; values ≥ 0.7°C were considered as thermal difference)	41%, 95% CI: 31-51
Lehmuskalio 1999 Finland	830 young male conscripts (mean age 19.3 years (range 17 - 29 years) divided into 4 groups by climatic home region on the south-north axis of Finland	Observational study (questionnaire) (LOE 4)	Risk of frostbite when using ointments	RR: 1.54, 95% CI: 1.17-2.03	No matched controls were used
Lehmuskalio 1995 Finland	913 young male conscripts with local frostbite of the head that needed medical attention and 2478 uninjured control conscripts	Observational study (LOE 4)	Risk of frostbite when using ointments	OR (cheeks): 3.3, 95% CI: 1.91-4.69; OR (ears): 4.5, 95% CI: 3.70-5.30; OR (nose): 5.6, 95% CI: 5.27-5.93	No matched controls were used; odds ratios were not adjusted for confounding variables; boundaries of the confidence intervals are own calculations assuming a normal distribution

Comment(s)

The study with the highest study design (intervention study; within subjects design) has measured an indirect outcome (skin temperature) (LOE 2b). Therefore, it was decided to include two studies with lower study design (observational studies) that has determined the risk of frostbite as an outcome (LOE 4).

Editor Comment

KMJ

Clinical Bottom Line

Evidence from one intervention study showed that using emollients resulted in statistically significantly more test persons with a decrease of facial skin temperature than with an increase of skin temperature. Evidence from two observational studies showed that the use of emollients resulted in a statistically significantly higher risk of frostbites when compared to not using emollients. In conclusion, evidence suggests that the use of emollients on the face is not a good protection against frostbites and would even result in a higher risk of frostbite. This evidence review and conclusion was part of an evidence-based practice guideline project, in which an expert panel also formulated practice experience and expert opinion.

References

1. Lehmuskallio E, Rintamäki H, Anttonen H Thermal effects of emollients on facial skin in the cold Acta Derm Venereol. 2000;80(3):203-7
2. Lehmuskallio E Cold Protecting Ointments and Frostbite. A Questionnaire Study of 830 Conscripts in Finland Acta Derm Venereol 1999; 79: 67-70
3. Lehmuskallio E, Lindholm H, Koskenvuo K, Sarna S, Friberg O, Viljanen A Frostbite of the face and ears: epidemiological study of risk factors in Finnish conscripts BMJ 1995;311:1661-3