In recent years, Germany has seen an increasing use of emergency rooms. For example, the number of cases in emergency rooms rose by approximately 34 percent between 2009 and 2015 (Sachverständigenrat zur Begutachtung der Entwicklung im Gesundheitswesen (SVR), 2017). This increase can be explained, on the one hand, by the age structure of the population and the multimorbidity of increasingly older people, and, on the other hand, by a change in the treatment spectrum of emergency rooms. Initial studies indicate that patients are increasingly visiting emergency rooms who would have received adequate treatment in an outpatient setting (Carret, Fassa & Domingues, 2009). This inadequate utilization results in overcrowded emergency rooms, long waiting times for patients, overworked healthcare personnel, and higher costs at higher levels of care (Haas et al., 2015; Köster, C., Wrede, S., Herrmann, T., Meyer, S., Willms, G., Broge, B., Szecsenyi, J., 2016). The situation is becoming more acute in Germany due to the peculiarities of the German healthcare system: Emergency care structures exist parallel in the outpatient and inpatient sectors. Coordination of the two care services across sectors does not yet occur regularly. The sector boundary between outpatient and inpatient care is suspected of being a cause of gaps in care and a lack of coordination. This can be attributed not least to the parallelism in planning, financing, and service provision, which is to be changed as part of the emergency care reform.

>> The associations of statutory health insurance physicians in Germany are currently responsible for ensuring a nationwide, contracted medical on-call service. These on-call services vary greatly from region to region, ranging from telephone on-call services to transport services and emergency rooms. In parallel to these care services, hospital emergency rooms exist. Patients often lack a clear understanding of the responsibilities within the healthcare system. They frequently lack information about the respective responsibilities and appropriate contact points (Somasundaram, Geissler, Leidel & Wrede, 2018). The choice of care area in cases outside of office hours is therefore often driven by the available information as well as their own perception of pain and urgency (see also Schmiedhofer, Searle, Slagman & Möckel, 2017). Patients incur no additional costs for using on-call services or emergency rooms. Patients are often tempted to use the best possible and most efficient service for them. Visiting an emergency room promises same-day treatment, prompt treatment, and comprehensive diagnostics (e.g., X-rays or computed tomography scans) regardless of specialist discipline (see also Schmiedhofer et al., 2017). A visit to the emergency room often offers an alternative in cases of waiting times for specialist appointments and diagnostics in the outpatient setting, as well as treatment outside of regular office hours, allowing for a visit outside of working hours.
Investigation of preventable cases
in emergency rooms
In a recent study by Schüttig and Sundmacher, an instrument was developed to identify potentially preventable cases in the emergency room, so-called outpatient-sensitive emergencies, in routine data (Schuettig & Sundmacher, 2019). Diagnoses were thus classified that could potentially also be treated by contracted physicians in standard care. This allows conclusions to be drawn about the extent of preventable cases in emergency rooms. The concept of outpatient-sensitive emergencies is based on that of outpatient-sensitive hospital cases (Purdy, Griffin, Salisbury, & Sharp, 2009; Sundmacher et al., 2015; Weissman, Gatsonis, & Epstein, 1992). The concepts of outpatient-sensitive emergencies and outpatient-sensitive hospital cases allow the investigation of potential implications of limited access or underutilization in the outpatient sector. In contrast to outpatient-sensitive hospital cases, outpatient-sensitive emergencies only consider cases in emergency rooms. Compared to inpatient hospitalizations, the analysis approach of outpatient-sensitive emergencies is limited to outpatient emergencies in the hospital.
The aim of the study was to estimate the number of cases treated in emergency rooms that could potentially also be treated on an outpatient basis by contracted physicians in standard care. For this purpose, a panel of 30 physicians was recruited, half of whom were outpatient and half inpatient physicians. The expert panel was surveyed in a multi-stage survey process to convert opinions into group consensus (Hasson, Keeney & McKenna, 2000). The Delphi method chosen is an established procedure that is also used, for example, in the consensus on guideline recommendations (Campbell, Braspenning, Hutchinson & Marshall, 2003). Based on this process, a list of outpatient-sensitive emergency diagnoses and reasons for emergency room visits for the selected symptoms and illnesses could be concretized. The physician panel was also asked whether the agreed-upon diagnosis could also be treated by the on-call service. Based on its experience, the expert panel also estimated the proportion of cases in a diagnostic group that could be treated through outpatient care and thus potentially avoided in the emergency room.
The results of the study allow for a retrospective review of care provision and the analysis of factors that favor emergency department use for potentially preventable cases. Outpatient-sensitive diagnoses can only be partially prevented through improved access and effective care (Sanderson & Dixon, 2000). Patient preferences, socioeconomics, experiences with the healthcare system, and other factors influence the decision to visit an emergency department (Boushy & Dubinsky, 1999). In regression models, the authors also examined factors that favored emergency department use for the identified outpatient-sensitive diagnoses. Supply-side factors of healthcare and patient characteristics were examined using spatial regression models.
Estimation of the extent of potential
avoidable cases in emergency rooms
The two-stage Delphi survey identified ICD diagnoses that were assessed by at least 65% of the participating physicians as treatable on an outpatient basis. The 135 selected diagnoses were categorized into 31 diagnoses according to the ICD classification.
Diagnosis groups were summarized (Table 1). These included, for example, general symptoms, back pain, and hypertension. The survey revealed significant differences in the assessment of the treatability of diagnoses by on-call services. For example, while 100% of respondents rated the diagnosis of acute nasopharyngitis as treatable on an outpatient basis, only 42% rated the diagnosis as treatable by the on-call service.
Expert assessments of the potential avoidability of cases in emergency rooms varied across the 31 diagnosis groups between 291 and 871 percent (Table 1). In 2015, 4.4 million patients with at least one outpatient-sensitive diagnosis were treated in an emergency room. Private health insurance (PKV) patients are not included in these analyses, nor are statutory health insurance (GKV) patients from the federal states of Bremen, Hesse, and Rhineland-Palatinate. However, since the coding of diagnoses in emergency rooms does not allow conclusions to be drawn about primary and secondary diagnoses, this total would overestimate the number of potentially avoidable cases. The respective diagnosis is not always the actual reason for treatment. The number of cases therefore represents a rough guide to the extent of potentially avoidable cases. Figure 1 shows the age and gender distribution of utilization in 2015. Younger patients in particular visited emergency rooms with outpatient-sensitive diagnoses. For both sexes, decreasing rates of outpatient-sensitive emergencies can be observed with increasing age up to the age group of 70 years and older. In addition, it can be observed that age groups with higher utilization of emergency departments with outpatient-sensitive diagnoses tended to have fewer outpatient physician contacts and tended to be less morbid. Regression models based on data at the district and urban district level examined additional factors: The influence of outpatient sector and patient characteristics on age- and gender-standardized rates of outpatient-sensitive cases in emergency departments was empirically estimated. The analyses show a negative correlation between the number of physician contacts and physician density in a region and the rate of outpatient-sensitive emergencies. The more frequently patients were treated as outpatients and potentially in continuous outpatient treatment, the less likely they were to visit the emergency department due to an outpatient-sensitive diagnosis.
gnosis. It was also shown that the higher the physician density in a region, the lower the rate of outpatient-sensitive emergencies. A negative correlation between voter turnout in a district or independent city and the rates of outpatient-sensitive emergencies was also confirmed. Voter turnout in a region can be interpreted as an indicator of social participation and attitudes toward public services.
The agreed outpatient-sensitive diagnoses were
in additional analyses
differentiated between general medical and surgical treatment spectrums to account for the special care situation in the case of injuries. The proportion of people with a migration background showed a significant positive influence on the overall rates and the rates of general medical diagnoses. For surgical diagnoses, however, this correlation was negative and not significant. This result may be due to the fact that cultural differences and familiarity with the local healthcare system play a role, particularly in the case of general medical symptoms.
have an impact on outpatient-sensitive emergencies.
Reasons for emergency room visits with outpatient-sensitive diagnoses
The study further identified patient motivations for emergency department visits with the agreed-upon diagnostic groups. Based on literature findings and interviews with emergency physicians, potential reasons for emergency department visits with outpatient-sensitive diagnoses were identified (Steffen, Tempka, & Klute, 2007; Uscher-Pines, Pines, Kellermann, Gillen, & Mehrotra, 2013). Seven motivations for emergency department visits were thus proposed to the expert panel:
• necessary diagnostics to exclude suspected diagnoses
• Comorbidities and overall patient condition
• Patients’ sense of emergency
• the patient's desire to visit the emergency room rather than the outpatient sector
• a referral by a practicing physician
• no competent outpatient doctors within easy reach
• lack of information from patients about responsibilities and contact points in the healthcare system

In addition to these reasons, the experts were able to add further reasons.
In the survey, patients' perception of an emergency was the most frequently cited reason for visiting emergency rooms. Patient preference for being cared for in the emergency room rather than in an outpatient setting was selected as the second most common reason for using emergency rooms for outpatient-sensitive diagnosis groups. The lack of nearby outpatient physician availability was cited as the main reason, particularly for skin diseases and psychological illnesses. Comorbidities and the overall condition of patients were selected by the experts as the most common reason for only one diagnosis group (diabetes mellitus). For the diagnosis group "open wounds and fractures," the need for diagnostics to rule out suspected diagnoses was the majority reason for visiting the emergency room. In summary, patient factors such as self-assessment, convenience, and the advantages of hospital treatment compared to outpatient treatment were identified as the main reasons for using emergency rooms for outpatient-sensitive diagnoses.
Studies by Schmiedhofer et al. (2016) and Somasundaram, Geissler, Leidel & Wrede (2018), which reached patients in emergency departments through face-to-face, open-ended interviews or questionnaires, identified similar reasons for patients' visits to the emergency department. The authors also cite lack of knowledge about treatment structures and sector boundaries, as well as concerns about health status, as common reasons for emergency department visits.
Conclusions, limitations of the study
and health policy relevance
Based on the results of the described study, potentially preventable cases in the emergency room can be classified and analyzed. Based on a Delphi consensus process with a panel of 30 physicians, an initial list of outpatient-sensitive diagnoses was compiled. This list does not represent a comprehensive list of potentially preventable diagnoses, but was based, as a first step, on diagnoses that are particularly frequently treated in emergency rooms. Based on this list, regional rates of outpatient-sensitive emergency rates can be calculated as a first indicator of regional differences and examined in further analyses. These differences in the use of emergency rooms can be identified based on routine data. Data are collected routinely and are not subject to any bias due to patient self-assessment.
The study results indicate that there is a correlation between access to outpatient care and physician retention, and the use of emergency rooms for outpatient-sensitive diagnoses. We find comparatively high rates of outpatient-sensitive emergencies, particularly among younger age groups with fewer physician contacts. This may be due to lower physician retention, particularly among younger age groups. Improved patient outcomes and avoided disruptions in care can be achieved, in particular, through continuous treatment. This is often not the case when emergency rooms are used.
The results of the regression analyses continue to show a positive correlation between the proportion of people with a migration background and rates of outpatient-sensitive emergencies. This can be explained by cultural differences in the use of emergency rooms and language barriers. However, the result may also indicate a lack of knowledge about the areas of responsibility in the German healthcare system. If this is the case, providing information brochures about suitable contact points and contacts with foreign language skills could be a suitable measure to address this problem. Among the general population, knowledge about the responsibilities and contact points in the healthcare system also appears to be insufficient.
The regression analyses also showed a negative association between voter turnout in a region and the outpatient-sensitive emergency rate. Since voter turnout in a region is often used as a proxy for social capital, it can be assumed that initiatives to navigate the healthcare system are particularly relevant in regions with low social capital. The expert panel also assumed that the patient preference to visit the emergency room rather than the outpatient sector is one of the most common reasons for using emergency rooms, which could be addressed through patient management. In this regard, the reform of emergency care can contribute to patients being better guided through an integrated system in the future. According to the current draft of the law to reform emergency care, the aim is, among other things, to create a common emergency control system and to set up integrated emergency centers in hospitals in order to better coordinate the different levels of care.
In regions with particularly high rates of outpatient-sensitive emergencies, it is necessary to examine whether timely access to the outpatient sector is not possible, whether long waiting times prevent outpatient care, and how these potential gaps in care could be addressed. The physician panel specifically focused on the lack of physicians in close proximity and waiting times in the outpatient sector for skin diseases and psychological illnesses. It is necessary to analyze whether the Appointment Service and Care Act represents an effective measure for providing patients with timely specialist appointments and thus resulting in changes in the emergency rates for these diagnoses.
The results of the expert survey revealed that patients' perception of an emergency was very frequently rated as the decisive reason for visiting the emergency room. Although participating physicians were able to provide multiple reasons, in most cases, necessary diagnostic testing to rule out suspected diagnoses was not selected as the reason for visiting the emergency room. This suggests that these diagnoses require higher-level patient management to direct them to the appropriate levels of care.
This study has limitations: The compiled list of outpatient-sensitive emergencies was based on the most common diagnoses in emergency rooms and on-call services, as well as other suggestions from the physician panel. The authors do not claim that the list of diagnoses is exhaustive. The diagnoses do not represent a definitive list of diagnoses that should not be treated in emergency rooms, but rather diagnoses for which outpatient treatment would have potentially been possible. Therefore, no conclusions can be drawn regarding the treatment of individual patients. The validity of the results depends significantly on the composition of the physician panel. Future studies could supplement the findings with a larger panel including a broader spectrum of medical specialties. Nevertheless, the outcome of consensus procedures always remains subjective and based on the experience and knowledge of the panel.
The reported study provides a basis for a retrospective analysis of emergency room utilization. Based on the outpatient-sensitive emergency rates, abnormalities can be identified and care patterns analyzed, which can justify measures to guide patients through the healthcare system in a needs-based manner.