Improving diagnosis in health care: laboratory medicine

Advancements in laboratory automation and their impact on diagnostics

Laboratory automation has made significant advancements in recent years (Table 2), which have improved the reliability of preanalytical processes [4]. It is now widely acknowledged that a substantial proportion of testing errors – estimated to be as high as 70% – originate from the preanalytical phase [5]. This phase encompasses all processes and activities preceding sample analysis, such as specimen collection, handling and preparation. Errors in this phase often stem from manually intensive tasks, including sample labeling, transport and processing [6]. Preanalytical errors, including errors in specimen collection, preparation, labelling and transport, are particularly common in patient care areas that are not under the direct management of clinical laboratories.

Modern laboratories now routinely integrate multiple levels of automation that address specimen check-in, preparation for testing (e.g., automated centrifugation for serum or plasma separation), and aliquoting for subsequent offline analysis or storage for future re-testing or storage in biobanks [7]. Automated sample verification can include hemolysis, icterus and lipemia (HIL) detection, as well as sensors to measure specimen volume and detect bubbles or clots [8]. Implementing these tools reduces the risk of human (mainly clerical) errors and enhances the standardization of preanalytical workflows, improving diagnostic accuracy and efficiency.

Laboratory automation has also been applied to analytical phase processes. For instance, modern laboratory analyzers incorporate sophisticated algorithms to identify anomalous analytical reactions in real-time, which can then trigger reflex or repeat testing based on predefined criteria, thereby ensuring result reliability and reducing the need for manual interventions [9]. Additionally, automated quality control (QC) management has emerged as a transformative feature, wherein analyzers can now schedule and perform QC assessments at fixed intervals, automatically flagging deviations and mitigating errors. Some recent analyzers also use reagent cassettes, including pre-configured internal QCs, eliminating the need for human intervention and enhancing reproducibility [10].

An important breakthrough in laboratory automation has been the development of automated sample storage and retrieval systems, which enable laboratories to efficiently manage vast repositories of biological specimens while providing real-time tracking and retrieval capabilities. This facilitates seamless sample re-evaluation, retrospective analysis and research, without compromising integrity or traceability [11]. Another advance in laboratory automation is the integration of multiple analytical platforms encompassing very different analytical techniques. While clinical chemistry and immunochemistry platforms have traditionally been consolidated, new technological frameworks allow the direct integration of genetic testing and liquid chromatography-mass spectrometry (LC-MS) analyzers within a total laboratory automation (TLA) model [12]. This integration enhances patient safety by reducing inter-system variability, minimizing manual handling errors, saving patient blood, reducing turnaround times and improving overall laboratory throughput, thus redefining modern laboratory operations and making them more efficient, accurate and scalable.

Advancements in molecular and genetic testing and their impact on patient safety

Molecular and genetic testing is revolutionizing healthcare by enabling earlier and more accurate diagnoses, personalized treatments, and improved disease prevention strategies (Table 3). Next-generation sequencing (NGS) allows for comprehensive genetic analysis, facilitating better characterization of various types of cancers and screening for rare diseases and hereditary conditions [13]. However, the implementation of NGS varies among clinical laboratories, and guidance and standards are evolving. For example, laboratories vary in their identification and reporting of potential disease-associated variants, requiring the need for clinicians who use these tests to have a clear understanding of limitations. Innovative technologies like Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR), especially CRISPR-Cas9 and Cas12/Cas13 systems, can be engineered for identifying specific genetic sequences and detecting pathogens or mutations in genetic material [14]. The test format can be easily adapted to fit portable diagnostic platforms (i.e., point-of-care instrumentation), reducing the need for specialized lab equipment [15]. Single-cell sequencing provides detailed insights into individual cell genetic makeup, enabling a better understanding of disease progression and accurate personalization of therapies [16].

Microfluidics technology is the foundation of Lab-on-a-Chip (LoC) systems, enabling the deployment of fast and highly accurate assays [17]. Miniaturized systems require less sample volume and can reduce errors, contamination and reagent waste [18]. These platforms can also support the simultaneous detection of multiple biomarkers, enabling more comprehensive and “green” diagnoses. This technology is also particularly suited for expanding access to diagnostics accessibility in underserved areas such as rural areas and low-income countries [19].

Liquid biopsy represents a promising and non-invasive method for detecting and monitoring cancer biomarkers, offering several practical and clinical advantages over traditional tissue biopsies [20]. Unlike tissue biopsies, which often require invasive procedures to collect specimens, liquid biopsies only need a simple blood draw, making the diagnostic process safer, faster, and more comfortable for patients while significantly lowering the risk of infection or complications. Liquid biopsies also play a crucial role in identifying minimal residual disease, detecting traces of cancer that may persist after treatment, and aiding in early relapse prediction [21]. As research and technological advancements continue, liquid biopsy is expected to become an essential tool in routine management of cancer patients, improving patient outcomes and advancing the era of personalized medicine. Another key advantage of liquid biopsies is their potential for real-time, continuous monitoring of disease progression. By providing dynamic insights into how a cancer evolves or responds to treatment, liquid biopsies may enable personalized and adaptive treatment strategies, detecting genetic mutations associated with resistance to targeted therapies, allowing oncologists to adjust treatment plans before resistance fully develops [22].

Pharmacogenomics, traditionally defined as the study of how an individual genetic makeup may influence response to certain drugs, can enable the tailoring of therapies based on genetic variations that influence drug metabolism, efficacy, and potential side effects [23]. In simple terms, pharmacogenomics refers to identifying genetic markers that can guide the selection of the most appropriate therapeutic agents and dosing regimens, thus minimizing the risk of side effects and optimizing treatment outcomes [24]. Clopidogrel, a widely prescribed antiplatelet medication for patients at risk of thrombotic events, is one of the many examples of the potential usefulness of pharmacogenomics in clinical practice [25]. The efficacy of clopidogrel is highly influenced by genetic polymorphisms in the CYP2C19 gene, which encodes the enzyme responsible for converting the drug into its active metabolite. Patients with genetic variations that impair CYP2C19 activity (i.e., “poor metabolizers”) may not achieve the expected therapeutic benefit from the standard dose of the drug [26].

Prevention of genetic disorders has also improved with carrier screening, which allows the identification of individuals at higher risk of transmitting severe hereditary conditions, aiding in informed reproductive choices [27]. Prenatal and newborn screening also allows early detection of genetic abnormalities, enabling the use of timely interventions and management to reduce the risk of developing more severe complications [28]. For example, with early diagnosis of Spinal Muscular Atrophy (SMA) and Severe Combined Immunodeficiency (SCID), early interventions can dramatically improve outcomes [29], [30]. For SMA, newborn screening is now part of routine practices worldwide, where a single capillary blood test can detect deletions or mutations in the SMN1 gene. When SMA is suspected, genetic testing can confirm the diagnosis and determine type and severity, enabling timely treatments (e.g., gene therapy, which provides functional copies of the SMN1 gene and prevents further motor neuron degeneration) [31]. In SCID, newborn screening involves testing for low T-cell receptor excision circles (TRECs), which reflects impaired T-cell production. Positive screening results prompt genetic testing to identify the specific mutation responsible for SCID, which can guide treatment, including bone marrow transplantation or gene therapy [32]. Early diagnosis of both conditions, made possible through newborn screening and genetic testing, significantly enhances the likelihood of successful treatment and better long-term health outcomes.

Infection control has been enhanced by rapid pathogen detection technologies [33] that aid in managing outbreaks, inside and outside the hospital environment, as early and accurate detection can significantly influence the outcome of containment measures and overall patient safety. For example, the global pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) highlighted the importance of rapid diagnostic testing. Real-time polymerase chain reaction (RT-PCR) tests, rapid antigen tests, and other diagnostic innovations allowed healthcare providers to quickly identify infected individuals, even before symptoms appeared [34]. This allows healthcare workers to isolate infected patients promptly, reducing transmission (including within healthcare facilities). Identifying pathogens responsible for hospital-acquired infections enables healthcare facilities to implement specific infection control practices, such as isolating infected patients and thoroughly disinfecting affected areas, preventing further spread within the hospital setting [35].

Rapid pathogen detection is critical in controlling multidrug-resistant organisms (MDROs) [36]. The World Health Organization (WHO) has long warned about the consequences of antimicrobial resistance (AMR), calling for immediate global action [37]. AMR makes infections more challenging to treat and leads to longer hospital stays, more intensive care and increased mortality [38]. Thus, it is essential to identify bacterial and fungal strains and determine whether they are resistant to specific antibiotics and antifungals [39].

Nonetheless, genetic testing requires robust ethical standards and regulatory frameworks to prevent commercializing tests with uncertain diagnostic utility, such as multi-cancer detection assays that claim to detect a broad spectrum of cancers. Both false-positive and false-negative results could lead to patient harm [40].

Advancements in artificial intelligence and machine learning and their impact on patient safety

AI, machine learning (ML), and digital pathology have the potential to profoundly transform various aspects of healthcare, including contributing to improved patient safety [41]. AI-driven tools, particularly deep learning models (DLMs), have been used to analyze medical images, pathology slides, clinical data, and laboratory results. In laboratory hematology, the adoption of AI-driven digitalized systems for blood cell classification has surged, making them an integral component of modern diagnostic instrumentation [42]. Automated blood smear production has standardized the process, minimizing inter-human variability and improving reproducibility [43]. AI-powered classification tools improve accuracy by detecting subtle morphological abnormalities that may be overlooked by human observers, facilitating the early detection of hematological disorders such as anemia, leukemia and infections. These systems also streamline laboratory workflows, reducing turnaround times and optimizing resource utilization [42].

AI may also improve laboratory data verification, validation and interpretation [44]. In data verification, AI-driven systems can be instructed to cross-check laboratory results in real time, identifying preanalytical errors (e.g., sample mislabeling, contamination, or inadequate volume) and flagging potential instrument malfunctions. For data validation, AI can compare longitudinal test results against patient history, identifying discrepancies between new results and prior patient records. AI-powered decision support systems can suggest repeat testing when results are incongruent with clinical expectations, ensuring that test outcomes align with disease patterns and reducing inaccuracies [45].

In laboratory medicine, AI has the potential to expand algorithmic test ordering. [46]. Laboratory diagnostics has long relied on standardized protocols and manual interpretation of results, and the integration of automated diagnostic algorithms within laboratory instrumentation or the laboratory information system (LIS) introduces a dynamic approach. Especially in high-throughput laboratories, AI-driven diagnostic algorithms can help optimize test scheduling and resource allocation, enabling faster turnaround time and improved operational efficiency. For example, consider the long-standing “TSH-reflex”. In traditional thyroid function testing, thyroid-stimulating hormone (TSH) is often used as the first-line screening test. According to the “TSH-reflex”, when TSH levels fall outside a predefined reference range, additional tests – such as free T4 and sometimes free T3 – are automatically performed to assess whether the abnormality underlines conditions like hypothyroidism or hyperthyroidism. This reflex testing approach reduces unnecessary assays, conserves time and biological samples, and enhances patient convenience by minimizing the need for additional blood draws if further testing is warranted based on initial results [47].

In terms of data interpretation, AI may enhance clinical insights by identifying early disease markers for faster and more efficient diagnoses, and correlating multiple biomarkers to refine differential diagnoses and supporting personalized medicine by integrating laboratory data with genetic, imaging and clinical information [48].

Challenges remain in the use and implementation of AI in clinical laboratory settings [49]. First, not all clinical laboratories are positioned to integrate AI into their workflow. Second, for those applications in which population-level data is required to inform result interpretation or application of a diagnostic algorithm, the dataset may be incomplete. Patient and family-level data may not be readily available, and population-level data is incomplete for persons of many ancestries. This does not preclude the use of AI, but rather argues for developing workflows that identify these limitations and alert users in a way that supports informed decisions.

Advancements in patient safety culture

Laboratory medicine has witnessed significant advancements in patient safety culture over the past decade, facilitated by developing and implementing robust reporting and learning systems.

Laboratory medicine has historically operated within a framework that often isolated its processes from other clinical practices. This siloed approach has sometimes led to communication gaps and errors that adversely affect patient care. After acknowledging these challenges, a concerted effort has been fostered to better integrate patient safety concepts and competencies into clinical laboratory science. Clinical laboratories have developed and implemented sophisticated reporting systems designed to capture, analyze and learn from errors [50]. These systems encourage healthcare professionals to report near misses and patient safety events, fostering a culture of transparency and continuous improvement. The International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) Model of Quality Indicators (MQI) in laboratory medicine, originally designed to monitor performance across different phases of laboratory testing [51], has been increasingly applied to more laboratories, standardizing error detection and recording, helping laboratories to identify the most important areas of improvement opportunities, and enabling benchmarking between institutions [52]. The Agency for Healthcare Research and Quality (AHRQ) has promoted Patient Safety Learning Laboratories (PSLLs). These laboratories adopt a multidisciplinary approach, using systems engineering to identify and address safety vulnerabilities. By fostering collaboration among healthcare providers, educators and policymakers, PSLLs aim to develop innovative solutions that enhance patient safety [53]. The Centers for Disease Control and Prevention (CDC) has introduced the “Core Elements of Hospital Diagnostic Excellence” to improve hospital diagnoses [54]. These elements include leadership commitment, multidisciplinary expertise, and continuous monitoring [55].

Another innovation that has gained momentum over the past several years is the concept of the diagnostic management team. The concept is based on the recognition that clinical laboratory expertise can provide invaluable insights to clinicians in diagnosis and patient management and, in particular, may reduce diagnostic errors attributable to laboratory testing [56]. These insights are most effectively incorporated into patient care when laboratory professionals are recognized as important members of the healthcare team and actively contribute to the care of the patient beyond providing a written or electronic copy of test results. This can include test selection, specimen collection, preparation, transport and result interpretation, all of which have been documented to be important sources of errors that affect patient safety [57]. Tumor boards and antimicrobial/antibiotic stewardship teams illustrate the diagnostic management team concept in that clinical laboratory professionals collaborate with clinicians on patient care decisions.