A survey of doctors reveals that few laboratory tests are of primary importance at the Emergency Department

Introduction

The clinical protocols at the Emergency Department (ED) of our hospital are set up so as to keep patients at the ED for as little time as possible [1]. The ED has a number of diagnostic tools to assist in deciding whether patients can be sent home or must be admitted for further treatment. In a recent, small-scale survey, we demonstrated that laboratory tests are among the ED’s most important diagnostic tools for medical decision making, in addition to anamnesis and physical examination [2]. Providing ED doctors with optimal facilities for ordering essential laboratory tests as efficiently as possible is therefore of major importance.

Approximately 36,000 patients are admitted to our ED each year. Although the majority of the acute patients reach the hospital via the ED, some patients circumvent the ED and are referred directly to a specific department. These include all patients with suspected heart problems (whether or not an ECG has been performed), who are sent directly to the Cardiac Care Unit, and pregnant women with pregnancy-related complaints, who are referred directly to the gynaecologists and delivery rooms.

Laboratory tests are requested for 40%–50% of the patients presenting at the ED. Doctors can request laboratory tests both as pre-defined test panels and as individual tests. Test panels, also called request protocols, form an integral part of clinical protocols [1]. Our ED currently offers 11 pre-defined test panels, which are largely disease oriented, e.g., ‘Lung’, ‘Abdomen’, ‘Sepsis’ and ‘Trauma’. Test panels improve quality and are convenient for requestors as they allow multiple tests to be ordered by means of a single mouse click. They also reduce the chance a doctor will forget to order a certain test, as all relevant tests are present in the panel. However, the use of test panels inevitably implies that not every test in the panel is genuinely needed for each patient; therefore, there will likely be some measure of overordering and overdiagnosis. Doctors, however, primarily focus on the quality of the diagnosis (and care), as well as speed and convenience in test ordering. They are typically less interested in and informed of the economic consequences of their activities. Therefore, the active involvement of laboratory specialists in defining new or modifying existing test panels results in major savings, even when limited to merely excluding redundant or unnecessary tests [1, 3, 4]. We recently illustrated this for the use of pregnancy tests by the ED [2]. Several other studies also suggest that uneconomical use of laboratory tests at the ED is common and that there is room for improvement in this regard [5–8].

The present configuration of our test panels evolved organically, in response to requests from ED doctors. As far as we know, however, no systematic survey or review of the laboratory testing needs at the ED has ever been conducted. The aim of the present study is to gain more insight into these needs and to examine whether the test panels currently available to doctors ought to be reorganised to promote efficiency.

Experimental procedures

Results

Discussion

Doctors involved with the ED report that clinical chemical laboratory tests are among the most relevant tools for the purposes of diagnosis, following patient history and physical examination [2]. As the ED’s aim is to diagnose patients as quickly as possible, the option of ordering laboratory tests in the form of test panels is faster and more convenient than ordering individual tests. In addition, the use of panels reduces the risk of forgetting specific tests, which could result in delay or misdiagnosis. On basis of our data, one or a few test panels for the ED could suffice to balance what is necessary from a medical perspective and efficient in logistical and economic terms.

Useful laboratory test panels have been proposed elsewhere, but these focused only on chemical tests [9]. It is acknowledged that only 49 doctors from one hospital (albeit a large one) responded to our questionnaire; however, we have no reason to think that the findings in other hospitals and among other professionals would be vastly different. Our data unambiguously show that only few laboratory tests are of major importance at the ED, and demonstrate that panels should contain haematological tests, including haemoglobin and leukocytes, in addition to chemical tests. Based on the results of this study, the test panels used at our ED have been adapted.

As noted, 35% of the responding doctors were employed at the ED, whereas the remaining doctors were from other medical departments. The relative distribution of responding doctors from the different departments correlates well with the distribution of doctors present at the ED on a regular day. Because doctors from the ED and doctors from other medical departments use the same electronic test ordering system, i.e., the ordering screens of which are designed specifically for the ED, our survey gives a good overall impression of doctors’ opinions on the relative importance of laboratory tests for patients presenting at the ED.

Questioning doctors in three different ways revealed that they see only a relatively small number of laboratory tests as decisive for most clinical decision making at the ED (Table 4). These included infection parameters (CRP and leukocytes) and kidney function tests (sodium, potassium, creatinin, and urea), and haemoglobin. We seriously considered the need to have urea as well as creatinin available as diagnostic test for kidney function. However, doctors at the ED indicated that they consider urea, together with the other kidney function parameters indispensable for evaluating the hydration status of newly admitted patients. Other tests appear to be less essential at the ED, although they may be important for patients with specific complaints; for instance, tests on blood gases, lactate, INR, D-dimer and urine screening. These tests better are not standard present in the panels, as is indeed the case in our hospital.

The doctors did not identify glucose as being of major importance at the ED. This is probably because, in our hospital, the majority of glucose measurements are performed at the ED using portable point-of-care devices (POCT). Only in specific situations glucose is measured in the laboratory, usually to confirm the result of a POCT or in non-urgent cases. Troponin was also largely absent in the responses, despite its central role in the evaluation of acute coronary syndrome. This may be explained by the fact that the majority of patients with cardiac symptoms are immediately referred to our Primary Cardiac Care Department. However, to increase generality of our conclusions, we tend to include both glucose and troponin among the indispensable laboratory tests for an ED.

Short turnaround times and high efficiency are crucial for the ED. Therefore, a useful approach may be to design a test panel in which all tests mentioned above are available on 24/7 basis. Given the nature of these tests, they indeed usually are available 24/7 in any hospital. Including them in a general panel has several advantages. Firstly, it allows for easy ordering of these tests for the majority of patients, increasing efficiency and safeguarding against the risk of essential tests being forgotten. And secondly, it ensures that the results of these essential tests are reported as quickly as possible, thus increasing the efficiency of patient diagnosis at the ED.

Although our study demonstrates that often only a few tests are necessary for diagnosing ED patients, a number of other tests are also important for more specific patients. Such tests include troponin, D-dimer, blood gases, coagulation status and blood type screening. If considered necessary, additional peripheral panels could be designed grouping these tests with others on the basis of patients’ major complaints. To optimise their setup and prevent redundancy, these panels should be developed in close collaboration between physicians and laboratory specialists, which will allow doctors to maintain their sense of oversight while also managing costs [1, 2, 4].

Although our study demonstrates that only few laboratory tests are of primary importance at the ED and that the use of smaller test panels could be an option, after extensive discussions with representatives of the ED it was decided not to introduce these smaller panels and to keep larger panels available. Doctors assume larger panels to prevent unnecessary delays at the ED, because all (possibly) essential tests are ordered at once and no tests will be forgotten. Incorporating essential tests in the test panels, potentially ordered by way of automated reflex testing, will also make the process more efficient and lean. Indeed, the costs of larger panels do not outweigh the costs for the hospital if patients have to stay at the ED longer. Doctors also feel that having too many different (small) panels available will increase the risk of ordering the wrong panels. Presented with the results of our survey, which show that many tests are often not needed for individual patients, doctors however agreed that the current test panels were in need of adaptation. An example is the redundant presence of liver enzymes in panels. To accommodate these insights, reflex testing was introduced in all ED test panels where applicable. For liver enzymes, only ALT is now present in the panels, and other liver enzymes will be measured only if the ALT result is abnormal. Similarly, fT4 and urea are now only measured if the TSH or creatinin results are abnormal, respectively. In the near future we will evaluate whether this approach successfully reduces redundant test ordering and lowers costs while preserving the logistical efficiency, quality and convenience required at the ED.