Understanding user requirements to improve adoption of influenza diagnostics in clinical care within Metro Manila

Abstract Background and aim Influenza diagnostics play a critical role informing in clinical management decisions and defining the global epidemiology of the disease to support public health responses. Use of influenza diagnostics within most low‐income and middle‐income countries remains limited, including in the Philippines, where they are currently used only for epidemiologic surveillance. The aim of this study was to define key considerations, including product characteristics, which may influence future adoption, uptake, and integration of influenza diagnostics into public and private clinical settings in this emerging Asian market. Methods Our study was conducted using a convenience sample of public and private hospital laboratories in Metro Manila. A usability assessment was conducted that included interviews with decision‐makers and direct observation of laboratory end users using 2 platforms representative of emerging diagnostic products: (1) a point‐of‐care antigen‐based rapid immunoassay diagnostic test paired with a reader and (2) a molecular diagnostic platform intended for decentralized use. Data were analyzed to assess user errors and device failure modes with each platform and to determine key considerations related to product adoption and uptake. Results The most difficult test step for most users on both platforms involved sample preparation. When deciding to adopt a new test, priority product attributes include performance, potential volume of demand from clinicians, equipment cost, and ease of use. Demand for new tests is likely going to be driven by clinicians, and policies and guidelines will be needed to support the introduction of new products. Conclusion Adoption of influenza diagnostics in Metro Manila is feasible but will require affordable products capable of satisfying needs for use in both epidemiologic surveillance and clinical management.


| INTRODUCTION
Globally, viral respiratory infections-including influenza-are a leading cause of morbidity and mortality. Early and accurate diagnosis of influenza can inform better treatment decisions, including reducing the inappropriate use of antibiotics. 1,2 Influenza diagnostics are not commonly used in the majority of low-income and middle-income countries (LMICs), and clinicians rely on signs as well as on a consideration of seasonal epidemics and locally circulating viruses, to inform case management and treatment decisions. However, as influenza may be difficult to differentiate from other acute febrile illnesses, the collection of an appropriate sample and an accurate laboratory diagnostic test is required to establish a definitive diagnosis. 3 While multiple options for influenza diagnostics exist, these tools vary considerably in accuracy, complexity, turnaround time, and other important performance characteristics. 4 In addition, the necessary performance requirements of the diagnostic will depend on the context of use, be it for public health surveillance or to inform patient care at a referral hospital. 5,6 Recent evidence suggests that inappropriate or insensitive diagnostic assays may lead to the mismanagement of influenza cases. 7 Moreover, reliance on central laboratories poses a challenge not only to improving patient case management but also to controlling nascent outbreaks and generating influenza surveillance data, particularly in places with limited laboratory capacity. This challenge is particularly evident in LMICs where laboratories have limited capacity and influenza epidemiology is poorly understood. 8 The influenza A (H1N1) pandemic of 2009 exposed the limitations of available diagnostic tools to support large-scale public health responses to influenza outbreaks. 9 This diagnostic gap affected the capacity of the international community to quickly detect and respond effectively to this emerging infectious disease, particularly during the early stage of the pandemic. While the poor performance of rapid immunoassay diagnostic tests (RIDTs) for influenza during the 2009 H1N1 pandemic is well documented, there is compelling evidence regarding the improved performance of new tests, particularly molecular assays, and their potential to provide clinical utility. [10][11][12][13][14][15] While many of the new and emerging diagnostic tests for influenza offer superior performance to RIDTs, it not yet clear if these tests will adequately meet the needs of the end user. These needs may include supporting clinical case management, public health surveillance for seasonal influenza viruses, or emergency responses to pandemics. 16 The cost, complexity, and turnaround times associated with these technologies often limit access outside of higher tier reference laboratories. 17,18 In both clinical case management and public health surveillance, as influenza diagnostics move closer to the patient and to the point of care (POC), the end user group also shifts, spurring a change in product design and development needs. It is critical to identify key product attributes that promote usability, as more complex diagnostics move into peripheral health laboratories and clinics. In addition to performance evaluations, usability assessments are needed to understand training requirements, whether target end users can successfully use the test and whether new tests can be feasibly integrated into health system policies and practices. This requires an assessment of nextgeneration diagnostic tools in target use cases.
The Philippines represents an ideal setting for evaluation, given the potential need for new influenza diagnostic products within its growing health care market that includes both private and public segments. Also, it is located in an area of global importance for influenza.
The Research Institute for Tropical Medicine (RITM) in the Philippines established an influenza surveillance program which includes sentinel sites in all regions across the country. Laboratory testing to support surveillance is conducted using respiratory specimens collected from patients identified at the peripheral health centers and outpatient departments of tertiary hospitals, that are then sent to RITM for testing via viral isolation techniques, and real-time reverse transcriptase polymerase chain reaction using the US Centers for Disease Control and Prevention protocol. 19 However, outside of RITM, routine influenza testing remains extremely limited in the Philippines. This is because, in part, of the lack of available diagnostic options with appropriate performance, cost, and user characteristics needed to expand the use of diagnostics for clinical care.
This study aimed to define key considerations, including product characteristics, which may influence future adoption, uptake, and integration of influenza diagnostics into public and private clinical settings in this emerging Asian market. Specifically, the study evaluated the usability and feasibility of introducing 2 influenza diagnostic platforms representative of emerging products-a POC RIDT paired with a reader and a moderately complex molecular diagnostic platform intended for decentralized use-among laboratory technicians in private and public health facilities. Manila area meet these criteria, and the hospitals included here were, then, selected based on proximity to RITM (Table 1).

| Participant characteristics
Six laboratory managers participated in this study. Laboratory technicians, rather than primary care providers, were the designated intended users for the tests. In the participating study hospitals, all diagnostic tests are conducted in the laboratory and at not at the bedside, regardless of complexity. Laboratory technicians often conduct rapid tests, such as dengue rapid tests. Forty-six medical technicians (medtechs) participated in the evaluation of the influenza tests. Common responsibilities of medtechs include sample receipt, sample processing and testing, and specimen collection, including blood extraction, and releasing test results.

| Study design
This study used mixed methods to evaluate the usability and explore the feasibility of integrating near-patient influenza diagnostic tests into clinical case management. To explore the feasibility of integrating nearpatient influenza diagnostic tests into clinical case management, key decision-makers in the management of the laboratory at each hospital were recruited, to participate in an interview using a structured questionnaire.
The questionnaire was piloted with the study team at RITM and allowed for "other" responses, to fully capture the participants' responses. The questionnaire focused on current testing procedures, laboratory systems and infrastructure, and how new tests are integrated into the laboratory.
Convenience sampling was then used to recruit and enroll laboratory technicians to participate in the usability testing of the influenza diagnostics. Sampling was based on availability and willingness to participate and provide informed consent. To evaluate the usability of current influenza  Run 1 and run 2 refer to fact that each user included in the analysis ran the test twice. The initial run is referred to as run 1, and the second run (run 2) was conducted at the same facility, 24 hours later.
platform technologies, end users were observed performing the procedure of 1 of 2 representative near-patient influenza diagnostic products intended for use at or near the POC ( Figure 2): To understand how near-patient influenza diagnostics could be integrated into the clinic flow, laboratory managers were asked to describe the current processes and practices for laboratory testing at the facilities.
Process maps (Figure 3) were generated based on their responses, which showed clear differences between inpatient and outpatient processes.
Some differences were observed between private and public facilities.
Notably, in outpatient private facilities, a cashier is involved to receive payment for the test. The cashier then uses the receipt of payment along with the physician's request to initiate the procedure, whereas in public or government facilities, no payment is required up front.
Currently, only the laboratory at RITM conducts confirmatory testing for influenza. Laboratory managers and the pathologist or head of the clinical laboratory are responsible for being aware of new technologies and also for making recommendations to the hospital administration regarding which tests should be added. They generally make a presentation and offer justification, including information on the cost of the investment as well as the equipment and training requirements.

| Form factor
Users appreciated kit components that were easy to handle, light, compact, and portable. Users noted that while the RDT reader was compact enough to be used at the patient bedside, bedside testing is not currently conducted at any of the participating hospitals. Although the molecular platform required additional hardware components, users familiar with PCR machines noted that the required equipment was much smaller than other molecular platform products. Users appreciated that all ancillary supplies needed to run both tests were readily available in their laboratories and easy to reorder.

| Workflow
The molecular platform required a lengthier and more complex workflow compared with the RIDT reader. However, users reported that the workflow was acceptable and suggested that once a technician was familiar with the test, it could be done quickly and with a shorter turnaround time compared with conventional PCR. This is in line with managers' notion of ease of performance as defined primarily by the amount of time needed for use and training. Users of the RIDT reader suggested that after repeat use, they would likely no longer need to refer to the instructions. As all users ran only 1 sample, some questioned how well the platforms would accommodate batch testing.
For both tests, a digital readout of the test result was preferred, as it provided a more definitive interpretation and increased confidence in the test result. A rapid turnaround time for results was considered a requirement for tests intended to inform patient care.

| Instructions for use
The availability of clear instructions with images and text was important, though preferences for digital or paper-based instructions were mixed.
Reported benefits of the digital IFU were the integrated timer and the fact that users did not need to move around paper instructions while handling test components. Drawbacks included a lack of familiarity with digital interfaces and the need to learn about a new test and a new IFU format at the same time. Users were clearly more familiar with paper instructions and required less intervention from the study team to help them move between steps. Users appreciated that both IFU relied on common words and short and direct sentences. Images reportedly helped users identify kit components and understand how they should be positioned. This evaluation had several key limitations. For one, this usability assessment did not include the use of actual specimens or specimen collection, which may impact usability and feasibility of adoption.

| DISCUSSION
Additionally, these platforms may require new systems and processes to accommodate the use of a potentially infectious specimen collected at the POC or near the patient. Furthermore, the adoption and use of new diagnostic platforms is highly contextual, and the findings of this study may not be applicable to other contexts, particularly in peripheral settings where access to health services and well-trained laboratorians are more limited. While the results of this study suggest that these tests could become part of routine practices with minimal training requirements, the medical technicians working in both settings were skilled, with several years of experience. This may or may not be the case in more peripheral health care settings with lower skilled medical technicians. This study sample is not exhaustive of all settings where better influenza diagnostics are needed.

| CONCLUSION
Although influenza diagnostics were not currently in use by any of the participating hospitals apart from RITM, results of the study indicated that the potential for influenza tests to be used within other laboratories in the Philippines is high. Priority attributes of future influenza diagnostic products should satisfy criteria considered important by laboratory managers: improved sensitivity and specificity over current rapid tests, high volume of tests demanded by clinicians, and low cost of equipment.
The impact of any improved diagnostic will be determined by whether the test is used for screening, differential diagnosis, or to inform treatment decisions, which will, in turn, be mediated by both official guidelines and clinical practice. Public and private clinical settings in Metro Manila require products that are accurate and affordable and which ideally can be used to support both clinical management and surveillance use cases. Integrating near-patient or POC influenza diagnostics into clinical care has the potential to not only improve data informing the global epidemiology of the disease but also strengthen clinical decision-making, including informing more appropriate antibiotic use.