PT - JOURNAL ARTICLE AU - Bergrath, Sebastian AU - Skorning, Max AU - Rörtgen, Daniel AU - Beckers, Stefan Kurt AU - Brokmann, Jörg Christian AU - Mutscher, Christina AU - Rossaint, Rolf TI - Is paper-based documentation in an emergency medical service adequate for retrospective scientific analysis? An evaluation of a physician-run service AID - 10.1136/emj.2009.086538 DP - 2011 Apr 01 TA - Emergency Medicine Journal PG - 320--324 VI - 28 IP - 4 4099 - http://emj.bmj.com/content/28/4/320.short 4100 - http://emj.bmj.com/content/28/4/320.full SO - Emerg Med J2011 Apr 01; 28 AB - Objective To investigate if paper-based documentation in the authors' emergency medical service (EMS) satisfies scientific requirements.Methods From 1 July 2007 to 28 February 2008, data from all paper-based protocols of a physician-run EMS in Aachen, Germany, were transferred to a SQL database (n=4815). Database queries were conducted after personal data had been anonymised. Documentation ratios of 11 individual parameters were analysed at two points in time (T1, scene; T2, arrival in emergency department). The calculability of the Mainz Emergency Evaluation Score (MEES, embracing seven vital parameters) was investigated. The calculability of the Revised Trauma Score (RTS) was also determined for all trauma patients (n=408). Fisher's exact test was used to compare differences in ratios at T1 versus T2.Results The documentation ratios of vital parameters ranged from 99.33% (Glasgow Coma Scale, T1) to 40.31% (respiratory rate, T2). The calculability of the MEES was poor (all missions: 28.31%, T1; 22.40%, T2; p<0.001). In missions that required cardiopulmonary resuscitation (n=87), the MEES was calculable in 9.20% of patients at T1 and 29.89% at T2 (p<0.001). In trauma missions, the RTS was calculable in 37.26% at T1 and 27.70% at T2 (p=0.004).Conclusions Documentation of vital parameters is carried out incompletely, and documentation of respiratory rate is particularly poor, making calculation of accepted emergency scores infeasible for a significant fraction of a given test population. The suitability of paper-based documentation is therefore limited. Electronic documentation that includes real-time plausibility checks might improve data quality. Further research is warranted.