Agreement between emergency room and discharge diagnoses in a population of injured inpatients: determinants and mortality.

BACKGROUND: Various factors contribute to the quality of care of an injured patient at the emergency room (ER), and a correct diagnosis can improve and accelerate care. The aim of this study was to evaluate the agreement between diagnoses assigned in the ER and those assigned after hospital admission to patients with unintentional injuries. We also tried to determine the factors that influenced the disagreement, and to evaluate if discordant diagnoses were associated with higher mortality risk. METHODS: All ER visits for unintentional injuries that were followed by hospital admission at the 60 emergency departments in the Lazio Region in 2000. Concordant diagnoses (ER/discharge) were established based on the Barell matrix cells. Logistic regression was used to assess the role of individual and ER care factors on the probability of concordance. A logistic regression was performed, where death within 30 days was the outcome and concordance was the determinant. RESULTS: We considered 22,892 ER visits for injury that were followed by hospital admission. In 62.2% of cases, the ER and discharge diagnoses were concordant. Higher concordance was found for older patients and less urgent cases. Factors influencing concordance were the hour of the visit, ER specialization degree, initial outcome, and length of hospital stay. Patients who had disconcordant diagnoses had a 30% higher probability of death. CONCLUSIONS: A correct diagnosis (i.e. confirmed at hospital discharge) at first contact with the emergency room is associated with lower mortality. Comparing administrative ER and hospital discharge data can be useful in emergency department management studies.

Does a pre-hospital emergency pathway improve early diagnosis and referral in suspected stroke patients?--Study protocol of a cluster randomised trial [ISRCTN41456865].

BACKGROUND: Early interventions proved to be able to improve prognosis in acute stroke patients. Prompt identification of symptoms, organised timely and efficient transportation towards appropriate facilities, become essential part of effective treatment. The implementation of an evidence based pre-hospital stroke care pathway may be a method for achieving the organizational standards required to grant appropriate care. We performed a systematic search for studies evaluating the effect of pre-hospital and emergency interventions for suspected stroke patients and we found that there seems to be only a few studies on the emergency field and none about implementation of clinical pathways. We will test the hypothesis that the adoption of emergency clinical pathway improves early diagnosis and referral in suspected stroke patients. We designed a cluster randomised controlled trial (C-RCT), the most powerful study design to assess the impact of complex interventions. The study was registered in the Current Controlled Trials Register: ISRCTN41456865--implementation of pre-hospital emergency pathway for stroke--a cluster randomised trial. METHODS/DESIGN: Two-arm cluster-randomised trial (C-RCT). 16 emergency services and 14 emergency rooms were randomised either to arm 1 (comprising a training module and administration of the guideline), or to arm 2 (no intervention, current practice). Arm 1 participants (152 physicians, 280 nurses, 50 drivers) attended an interactive two sessions course with continuous medical education CME credits on the contents of the clinical pathway. We estimated that around 750 patients will be met by the services in the 6 months of observation. This duration allows recruiting a sample of patients sufficient to observe a 30% improvement in the proportion of appropriate diagnoses. Data collection will be performed using current information systems. Process outcomes will be measured at the cluster level six months after the intervention. We will assess the guideline recommendations for emergency and pre-hospital stroke management relative to: 1) promptness of interventions for hyperacute ischaemic stroke; 2) promptness of interventions for hyperacute haemorrhagic stroke 3) appropriate diagnosis. Outcomes will be expressed as proportions of patients with a positive CT for ischaemic stroke and symptoms onset < or = 6 hour admitted to the stroke unit. DISCUSSION: The fields in which this trial will play are usually neglected by randomised controlled trial (RCT). We have chosen the cluster-randomised controlled trial (C-RCT) to address the issues of contamination, adherence to real practice, and community dimension of the intervention, with a complex definition of clusters and an extensive use of routine data to collect the outcomes.