Towards harmonization of quality indicators in laboratory medicine

Introduction

Accurate and efficient clinical laboratory testing is a critical component of high-quality patient care as laboratory test results influence most medical decisions, including diagnosis, prognosis, risk and predictive assessment, and prevention, screening and the monitoring of treatments and therapies. In addition, aggregate test result data are used for public health surveillance, healthcare performance measurement, and quality improvement [1]. The quality of laboratory testing, therefore, may greatly affect the quality and affordability of patient care and any defects or errors impact on the care of each patient as well as the costs incurred by the healthcare system [2]. However, the laboratory testing process is complex and consists of many steps, beginning with test selection and request, followed by sample collection, transport, processing and analysis, and finally, result reporting and interpretation. In 1975, on describing the testing process as the “brain-to-brain loop”, George D. Lundberg stressed that what counts is the final step of the cycle: the appropriateness of the action(s) undertaken on the individual patient and based on the laboratory information provided. Lundberg also stated “unless the appropriate action occurs, it is as if the cycle had never begun and is, at the most, a tragedy and, at the least, a waste” [3]. According to this viewpoint, which is consistent with a patient-centered approach, all steps of the testing cycle should be evaluated and continuously monitored to assure that high quality laboratory information is provided and that, in turn, an appropriate action is taken for the individual patient.

A body of recent evidence highlights that most errors in the “brain-to-brain loop” fall into neither the analytical phase nor the pre- and post-analytical steps under the control and/or jurisdiction of laboratory professionals [2, 4, 5]. In the last few decades, improvement in the reliability and standardization of analytical techniques, reagents and instrumentation, as well as advancement in information technology, quality control and assurance methodologies have decreased the analytical error rate by several-fold. More recently, thanks to the introduction of pre-analytical workstations and integrated pre-analytical systems, there has been a significant reduction in errors due to specimen preparation, centrifugation, aliquoting, pipetting and sorting [2]. Likewise, significant improvements have been achieved in the post-analytic phase, especially in data transcription, largely due to interfacing analyzers and laboratory information systems (LISs). Information technologies have also allowed a more rapid and effective validation of laboratory results and improvement in the timeliness of result notification, namely the notification of critical values [6, 7]. Therefore, intra-mural laboratory procedures are increasingly being considered safer than ever. However, on investigating the beginning and the end of the loop, it emerges that the pre-pre-analytical steps (initial procedures performed neither in the clinical laboratory nor, at least in part, under the control of laboratory personnel), and the post-post-analytical steps (final procedures performed outside the laboratory, consisting of receiving, reading, interpreting and utilizing the laboratory information for patient management) are more error-prone than ever. These activities are poorly evaluated and monitored, often because the process owner is unidentified and the responsibility falls in the “no-man’s land” between laboratory and clinical departments. Data from different clinical settings, such as primary care, internal medicine and emergency departments clearly show that the rates of errors in test request and result interpretation is unacceptably high, and translate into missed, delayed or erroneous diagnoses [5]: the risk of errors and patient harm in the “brain-to-brain loop” has been significantly reduced in processes developed in the laboratory, but it has remained quite high at the beginning and at the end of the loop, which lies mainly outside the traditional laboratory environment. For patients and for laboratory medicine, the relatively high rates of errors in the pre-pre and in the post-post-analytical steps calls for substantial reorganization, as well as the delivery of laboratory services through teamwork and interdisciplinary cooperation because most of the vulnerable steps are not and cannot be brought back under the direct control of laboratory personnel.

Quality indicators in laboratory medicine and harmonization

It has been demonstrated that performance and outcome measures can improve the quality of patient care. These measures support accountability and enable the comparison over time between providers, the effectiveness of delivered services and the improvement in patient safety being evaluated through the development and monitoring of specific indicators [8]. The dramatic decrease in analytical errors achieved in laboratory medicine is due, at least in part, to the fact that quality indicators (QIs) and quality specifications have been developed and introduced for the effective management of analytical procedures. Clinical laboratories can now measure, monitor and improve their analytic performances over time thanks to internal quality control rules, objective analytical quality specifications, and Proficiency Testing (PT)/External Quality Assessment (EQA) programs, which have provided clinical laboratories with a valuable benchmark based on objective data. The identification of reliable QIs in the total testing process is therefore a key step in enabling users to quantify the quality of laboratory services, but the current lack of attention to extra-laboratory factors is in stark contrast with the numerous studies on the multitude of errors that continue to occur in the pre- and post-analytical phase. Unfortunately, while some interesting programs on indicators of the extra-analytical phases have been developed in some countries, there is no consensus for the production of joint recommendations focusing on the adoption of universal QIs and common terminology in the total testing process. Since a variety of QIs and terminologies are currently used, there is a need to harmonize proposed QIs that must be: 1) patient-centered; 2) consistent with the requirements of the International Standard for medical laboratories accreditation (ISO 15189: 2007); and 3) conducive to addressing all stages of the total testing process (TTP). In addition, the process of harmonization of QIs consists of two compulsory steps: the identification of common QIs and a standardized reporting system. Essential pre-requisites of QIs appear to be: 1) importance and applicability to a wide range of clinical laboratories at an international level; 2) scientific soundness with a focus on areas of great importance for quality in laboratory medicine; 3) feasibility, both regarding the data availability and the definition of thresholds for acceptable performance; and 4) timeliness and possible utilization as a measure of laboratory improvement. Difficulties in identifying, defining, and ultimately implementing indicators that cover the various stages of the total testing process have been underlined by Shahangian and Snyder, who examined the peer-reviewed publications from January 1990 through July 2008 and identified 14 QIs that met two inclusion criteria: 1) the use of a quantitative measure associated with laboratory testing; and 2) the potential to be related to at least one IOM healthcare domain [9]. Other experiences described in the recent literature concern the use of QIs in Laboratory Medicine, namely those by a working group of the Catalonian Health Institute (ICS) [10–12], by the Brazilian Society of Clinical Pathology/Laboratory Medicine (SBPC/ML) [13] and by the Quality Assurance Scientific and Education Committee (QASEC) of the Royal College of Pathologists of Australasia (RCPA), supported by the Australian Government Department of Health and Ageing, which launched the project “Key Incident Monitoring & Management Systems (KIMMS)” [14]. While there are differences between these programs, they all contain the same common pre-analytical QIs, which should be grouped into two main categories for the evaluation of: 1) identification problems (unlabeled and mislabeled samples or samples suspected to be from the wrong patient); and 2) sample problems (clotted, hemolyzed, insufficient, contaminated samples), respectively. In the post-analytical phase, QIs address troubles in turnaround time and report transmission. However, according to the reappraisal of the “brain-to-brain concept” [15] and the need to cover all steps of the process [16–19] it appears important to include QIs on the appropriateness of test request and clinical advice in the pre-analytical phase. The same principle applies to the post-analytical phase in which some QIs should take into account issues, such as interpretative advice and critical results communication.

The pathway to the identification of harmonized QIs

The starting point for identifying reliable QIs is the identification of critical activities in the TTP as well as the analysis of the specific laboratory context, described in previous papers [9, 20]. This step seems easier to take if we consider the data in scientific literature highlighting the vulnerability of extra-analytical phases [2, 3, 21, 22].

Unfortunately, the monitoring of critical activities, which calls for more than one measure, should be interpreted in different ways. For example, while patient identification is without doubt a critical issue, as stressed in many recommendations made by accreditation and safety agencie event to keep under control can be affected by different organizational procedures depending on the specific laboratory as well as the laboratory staffs’ awareness of the patient’s centeredness. Consequently, different measures and incomparable data may originate from different laboratories. Collected data may refer to errors: 1) concerning identification of patient when the patient’s identity is not assured; 2) concerning identification of patient when the patient’s identity is assured; 3) detected and corrected before a test result is released; 4) detected and corrected after a test result is released; and 5) concerning the interchange of patients – in this case, the error can be counted once (if the exchange error is counted) or twice (if the number of patients that have been impacted is counted). The error rates concerning patient identification can thus be affected by different interpretations and measure that each laboratory should carry out.

It is therefore apparent that the path towards the harmonization of Quality Indicators in Laboratory Medicine requires not only common QIs, but also a relevant and valuable reporting system, the efficacy of QIs being closely linked to the reporting system designed; this calls for clear definitions of what is to be reported, who has to collect the information, when and how to report it, so that comparable data at a national and international levels can be obtained. The data reporting procedure must be straightforward, and easy to understand and use.

In order to comply with the above requirement, a Model of Quality Indicators (MQI) has been consensually developed by a group of clinical laboratories in a project launched by a working group of the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC); it aims to set up a user-friendly reporting system for laboratories worldwide and to allow the collection of comparable data from different laboratories, and its end aim is to provide a harmonized structure that assures the provision of timely and valuable information on the efficiency and efficacy of all processes, particularly those significantly affecting patient safety. The QIs included in the model have been identified taking into account the different organizational procedures carried out in clinical laboratories worldwide, and their related needs. Moreover, the data collection method has been developed consensually in order to assure consistency and transferability of data. Examples of the QIs proposed for patient identification are shown in Table 1; data collection and reporting is made easier for clinical laboratories if the problem (patient identification) is split into different QIs.

Table 1

Model of quality indicators: example of a quality indicator.

MQI and quality indicators

Currently, the MQI includes 56 QIs related to key processes (34 pre-, 7 intra- and 15 post-analytical phase) and three to support processes (Tables 2–5). An “experimental phase” was tested under real conditions by involved laboratories in order to check the feasibility of the system and to determine the kind of feedback that would be useful for acceptable to participating laboratories. The “experimental phase”, which is closed, has allowed improvement to be made to the design of the system, identification of the most appropriate possible data collection procedure and the ways in which the reporting system could be of the greatest possible benefit.

During the “working phase”, which is in progress and aims to collect data and to identify quality specification for all QIs, regularly issued confidential reports are provided to show the results of each laboratory and the statistical findings obtained by data processing. An evaluation of the results using the sigma metric method is included (Figures 1–3); as clearly shown in Figure 1, an improvement in the sigma values has been achieved over time. A detailed report on the sigma data for the same QI is shown in Figure 2, while Figure 3 reports the distribution of the sigma values among participating laboratories. Before issuing the report, collected data are continuously checked, in order to ascertain whether the collected data are reliable and to guarantee the validity of information provided in the report.

Figure 1

Laboratory report showing the sigma trend concerning the samples with inadequate sample-anticoagulant volume ratios.

Figure 2

Example of a laboratory report.

The data reported on the left are the laboratory values in percentage (%) and as sigma, respectively. On the right side the data reported are the mean of the sigma for the specific group of laboratories (e.g., Italian or Chinese) and the overall mean, respectively.

Figure 3

Example of a laboratory report: sigma distribution on the percentage of hemolyzed samples for chemistry tests.

The question as to whether or not it is opportune to include in the MQI many indicators is still open. On one hand, a Model with many QIs gives laboratories a choice, thus allowing them to use only those that are more suitable for their quality system. On the other hand, it might be discouraging, laboratories being overwhelmed by the effort and time needed to collect data. Importantly, however, the Model is a reporting system that should guarantee standardized data collection, and some indicators (e.g., patient identification) are part of a single event kept under control. Consequently, a greater number of QIs could help the laboratory to monitor the specific activities and facilitate the identification of subsequent measures for improvement.

Conclusions

According to the definition of the Clinical and Laboratory Standards Institute, the term “harmonization” refers to “the process of recognizing, understanding and explaining differences while taking steps to achieve uniformity of results such that different groups can use the data obtained from assays interchangeably” [23]. However, the scope of harmonization in laboratory medicine is more far-reaching than method harmonization and should cover a wider range of topics, namely all steps of the “brain-to-brain loop”, currently also referred to as the “end-to-end laboratory testing process” [1]. Some projects aiming to optimize laboratory services through global harmonization activities involving test names, reporting units and reference intervals are underway [24]. However, the road map towards global harmonization should take into consideration the current evidence of errors in laboratory medicine and the need to reduce the risk of errors not only within the laboratory walls but also in the pre-pre and post-post-analytic processes from a patient-centered perspective while involving all appropriate stakeholders right from the start [25]. In fact, most pre- and post-analytic laboratory errors involve some breakdown in the process, and both laboratory and clinicians bear mutual responsibility for these errors and for developing safeguards that will prevent them.

The value of QIs lies in their being a tool enabling users to quantify the quality of a selected aspect of care by comparing it with a highlighted criterion [8]. QIs should meet specific requirements: they must be evidence-based and implemented in a consistent and comparable manner across settings and over time. As pointed out, it is a great challenge to identify, define and finally use indicators that cover the various stages of the total laboratory testing process because the numerous processes involved concern different care operators in both the laboratory and in the clinical setting. The current independent and somewhat inharmonious outburst of national initiatives, on heterogeneous sets of QIs and related terminology may, on the one hand, be welcome, but, on the other, they are testimony to the need to reorganize and possibly unify ongoing projects [5]. The aim of the IFCC LEPS (laboratory errors and patient safety) project is to provide a common framework and to consensually establish a set of QIs that should cover all steps of the TTP with related quality specifications obtained by collecting data from several laboratories around the world. The main goals of the study design, grounded on international consensus achieved among professionals from clinical laboratories, are to identify reliable QIs and easy data collection. The development and availability of a specific website facilitate data collection and reporting, ensuring confidentiality of information and obviate costs for participating laboratories, all of which are valuable pre-requisites for any further development of the project. The search for quality and patient safety in laboratory medicine was initiated long ago, but there is still a long way to go: the journey must be taken step by step over time, and the harmonization of QIs is a milestone in this journey, as it will be the key to identifying, preventing and monitoring all procedures and processes at risk of error in the total testing process.

In conclusion in a global perspective, the concept of harmonization by definition entails the harmonization of quality assurance procedures. The use of a Model of Quality Indicators focusing on all phases of the TTP with related quality specifications will enable laboratories at home and abroad, to share their experience, learn by comparing themselves with others, and continuously improve their performance.

Conflict of interest statement

Authors’ conflict of interest disclosure: The authors stated that there are no conflicts of interest regarding the publication of this article.

Research funding: None declared.

Employment or leadership: None declared.

Honorarium: None declared.