NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Morel C, McClure L, Edwards S, et al., editors. Ensuring innovation in diagnostics for bacterial infection: Implications for policy [Internet]. Copenhagen (Denmark): European Observatory on Health Systems and Policies; 2016. (Observatory Studies Series, No. 44.)
Ensuring innovation in diagnostics for bacterial infection: Implications for policy [Internet].
Show details8.1. Introduction: complexity in demand expression
In the case of some technologies, and indeed diagnostics in particular, true demand – defined here as the technologies most needed or desired to improve patient outcomes – may not be expressed due to a multitude of factors. First, health care is typically provided within a complex organizational structure that influences demand. For example, the governance arrangements for deciding to adopt a new POC diagnostic often include many players. In hospitals, there is typically a team of people involved in decision-making – a mix of clinicians and laboratory staff, with the involvement of experts from the clinical specialties that will make use of the device, for example consultants from emergency medicine and intensive care. Other departments that may be involved include the pharmacy and staff involved in the maintenance of equipment. In some hospitals there is a dedicated POC testing manager who can help guide the decision-making process. These decisions are not made by the patient but rather on his or her behalf via agency relationships. Further, the demand by a clinician, for example, may be expressed by those who are unaware of the price of a particular technology. Additionally, the device is usually ultimately covered by payment by a third-party organization. Demand is also influenced by other factors, such as ethics, altruism or other financial and non-financial incentives in the health system.1 These complexities are also compounded by the difficulties in building a strong evidence base surrounding the relative merits of these technologies that includes their respective cost–effectiveness for clinical units, for hospitals and for society more broadly given their potential to help slow resistance to antibiotics.
From the perspective of diagnostics developers, these complexities can give them an altered view of demand and the willingness to pay for the technologies that meet it. Fragmented decision-making in many health care markets makes it extremely difficult for companies to understand the requirements of all key stakeholders. To be selected for use, a device might have to be approved by a national or regional authority, selected by a health care provider, specified by a particular clinical team and then chosen by doctors, often in consultation with patients. Finally, it may be the patients’ own reactions to the device that define its success in use.2 Ultimately, the basic task of identifying the customer can be complex. Indeed, developers need to understand the needs of the appropriate staff using and interpreting POCTs. Indeed, there are numerous examples of technologies being developed that lack the necessary characteristics for clinical adoption. One interviewee emphasized the over-reliance of the developers of a sepsis-related technology on microbiologists rather than physicians. The lack of understanding of “real-life” clinical decision-making has effectively rendered the device useless in many settings in which it was most needed. Another example of a good technology that failed to fit into the wider clinical context comes from a study into whether carrying out a POC blood test at a patient’s bedside would reduce the LOS in an emergency room setting. The POCT that was used only provided a limited biochemistry profile. While POC testing delivered quicker test results, the researchers found that patients were not leaving the hospital any faster because of the need to wait for additional necessary tests to be analysed by a central laboratory.3
A number of developers and industry consultants have raised the issue of divergent views among the microbiologists’ community as a challenge to gauging the priorities and focus of end-users of their products: even at institutions facing similar challenges, views on both pathology priorities, and desired solutions are often different. This adds another dimension to the fragmentation in the marketplace, and means diagnostic manufacturers need to engage with a wide range of professionals in the microbiology community in order to fully understand the potential demand for a given product.
8.2. Engagement to improve developer understanding of demand
The only way that developers can sufficiently understand true demand in order to produce a diagnostic capable of altering the patient care pathway and improving health outcomes is to work very closely with clinical staff and patients. There is evidence that a significant gap exists between the views of frontline clinicians and industry professionals on what constitutes an ideal POCT. In a recently published study by academics from Johns Hopkins University, diagnostic users and developers were surveyed about perceived barriers to using POCTs for STIs.4 The industry representatives identified problems such as complexity, unreliability and difficulty reading test results, whereas the clinicians placed much greater weight on workflow factors, such as the time frame of a test and how well it integrates into existing work processes.
This disparity in perceptions may, in part, explain why there are still POC diagnostics being marketed with characteristics that are potentially major barriers to adoption in practice. In an online survey of STI experts and professionals, barriers to routine use of POC diagnostics that were cited included tests that could only be undertaken in a laboratory and tests that interrupted work flow5. There are also examples of products reaching the market without sufficient preparation being made for their adoption; for example, when the breast cancer drug, Herceptin was initially approved for use in the United Kingdom, consideration had not been given to how the companion diagnostic, which is used to establish clinical eligibility for treatment, would be funded.6
At a conference at the University of Oxford in 2011 that brought together industry professionals, clinicians, academics, and regulators, a wide range of benefits were identified that could be achieved through early engagement.7 These included supporting prioritization of the development of new tests, increasing understanding of market trends that may influence a test’s future uptake and, through confirming unmet needs, improving the evidence base used to justify investment in R&D to investors. Engagement can also facilitate an understanding of how a new test will change existing care pathways and how it will change resource utilization – for example whether it could enable disinvestment in existing processes. As clinical needs and barriers to using a test vary depending on the clinical setting as well as contextual factors, such as the availability of pathology services,8 the approach to engagement needs to reflect this heterogeneity.
At present, although there is engagement between industry and clinicians, it is often informal. Carol Cresswell, POCT Manager for Newcastle Hospitals in England, reported that her team’s primary engagement channel with industry is visits from industry representatives, particularly from larger companies. This provides an opportunity to offer views on areas of unmet need and make suggestions on enhancements to devices; however, she felt that this was not ideal and was keen to see a move towards a more structured approach. The type of staff that the industry is in routine contact with, for example through providing sales, training and support for their products, are not always the frontline clinicians who will be using tests. Future users of a proposed test are best placed to identify problems with current processes and the potential impact of introducing a new device. Another challenge cited by Doris-Ann Williams MBE, Chief Executive of BIVDA is that compared to the pharmaceutical industry, diagnostic companies tend to have a much smaller sales force and so may have less opportunity for day-to-day contact with clinicians.
Dr Gary Thorpe, Biochemistry Director at the University of Birmingham believed that when industry seeks to work jointly with health professionals, there is sometimes unease among clinicians and suspicion about industry’s motives. This is a potential barrier to improving engagement. A physician from the United States interviewed for this study reported that if they experienced a problem with a diagnostic test, they were more likely to contact a regulatory body than pick up the phone to report the problem to the manufacturer. When questioned further, their view was that they simply did not trust the manufacturer. In the United Kingdom, there has been recognition at the highest levels within the NHS that suspicion such as this can hinder partnership working;9 there is a need for both the NHS and industry to work towards a more collaborative culture, for example through improving understanding of the benefits to society of a more joined-up approach.
While patients may not have as good an understanding as clinicians of the complexities of care pathways, they can offer an important insight into the patient experience – for example the impact of delays in receiving test results in outpatient care. To date, there has been limited exploration of patient acceptability of POC testing,10 but emerging evidence suggests that patient preferences can influence the uptake of a new diagnostic. For example, in a study evaluating the effectiveness of the BioStar Chlamydia OIA POCT, 6.8% of female adolescents tested at a public clinic in Atlanta were unwilling to wait 20 minutes for the results.11 A concern raised by Dr Ron Daniels, Executive Director at Global Sepsis Alliance, was that where engagement is narrow, individual or professional agendas may shift the focus of debate away from patients. He sees involving patient groups as a way to balance this effect and ensure that tests met patient needs.
To improve engagement, a number of national strategic initiatives have been established. In England, NHS organizations are being given the opportunity to bid for funding to host Diagnostic Evidence Co-operatives (DECs), which are designed to facilitate collaborative working between health professionals, the diagnostics industry, providers of NHS pathology services, academia and patient representatives, as well as support the generation of real-world evidence of the clinical utility and cost–effectiveness of devices.a In the United States, the National Institute of Biomedical Imaging and Bioengineering has supported the development of the POC Technologies Research Network (POCTRN).12 Institutions that have been designated POCTRN centres offer practical support to industry, such as needs assessment to inform device design and evaluation of the potential clinical impact of prototypes.13
8.3. Determinants of and barriers to uptake of new POC diagnostics
In 2001, the United States Institute of Medicine voiced the concern that science and technology are advancing more rapidly than health systems can consistently deliver them.14 In their ground-breaking report, Crossing the quality chasm: A new health system for the 21st century, the Institute identified a range of shortcomings, including constraints in exploiting the revolution in information technology and poorly organized delivery systems, both of which have significance in the context of the introduction and uptake of diagnostics in health care. In 2000, the Center for Health Care Quality at the University of Missouri estimated that the time lag between research identifying more effective treatment options and their adoption in practice was approximately 17 years.15 In the United Kingdom, the 2002 Wanless Report cited the United Kingdom as a “late and slow adopter of medical technology”16 and, more recently, in a report into how the adoption of innovation could be accelerated in England, the Department of Health singled out diagnostics as a key area for action.17
Studies on health technology adoption rates in the NHS in the United Kingdom have illustrated how the diffusion rate is dependent on the nature of the technology. While the cholesterol-lowering drug, simvastatin was adopted very rapidly, it took six years from the launch of coronary stents to rapid diffusion across the United Kingdom and a further two years before they were in widespread use. In the case of MRI scanners, only 70% of hospitals had access to the technology 17 years after it became available.18 More recently, it has been estimated that it takes around 10 years for widespread adoption of a new diagnostic test within the NHS.19
8.3.1. Quality control arrangements
Training
By design, POC devices are often simple to operate, but the potential consequences of inadequate training, including delayed or incorrect diagnosis, are significant. Real-life anecdotes shared by laboratory staff20 include POC glucose tests appearing abnormally high where the test strip has been contaminated by the patient or health professional not washing their hands, pregnancy tests being misinterpreted because the faint indicator line was not detected in the poorly lit sluice room where the tests were being interpreted, and the over-referral of patients to a specialist endocrinologist because the device used on wards to monitor patient urine chemistry was changed and staff misinterpreted readings from the new device. A small study across three hospitals in Northern Ireland in 2011 provided evidence to suggest that the quantity of user errors may be significantly higher for POCTs compared to central laboratory testing.21 Many of the errors identified in the study could have been avoided through better training or improved adherence to standard operating procedures.
To help ensure that users are competent to use tests, many countries require training either as best practice or as a regulatory obligation. In the United States, under the CLIA of 1988, all facilities examining human specimens for diagnosis must register with the CMS and obtain CLIA certification. However, the nature of registration and subsequently the training requirements vary according to the nature of the diagnostic device being used.22 In a number of circumstances, the FDA has waived tests from regulatory oversight, including where they are for home use, where the test’s simplicity makes it unlikely that erroneous results will be generated and where there is no reasonable risk of harm if tests are incorrectly undertaken. Where a provider is only offering waived tests, they can choose to apply for a “certificate of waiver”; this exempts them from routine inspections but they must adhere to the manufacturer’s instructions for performing the test and best practice still applies.23 In 2005, to support facilities, the CDC and the Division of Laboratory Science and Standards jointly published guidance on “Good laboratory practices for waived testing sites”.24 This recommends that careful consideration is given to ensure adequacy of training, including evaluating competence before staff perform training; ensure competence is maintained, particularly where testing volumes are low and making provision for turnover of staff. The guidance also advises that training is documented and suggests key areas that should be covered, including safety and quality control procedures.
Where a facility intends to use tests that have not been waived from regulatory oversight, they need CLIA certification and to comply with a range of regulatory requirements; on training this includes ensuring that, prior to testing, staff have education and experience relevant to the type and complexity of services offered and have demonstrated their competence, and that policies are in place to assure continued competence. For facilities using waived tests, an alternative to applying for a CLIA Certificate of Waiver is establishing an agreement to work under an existing laboratory with CLIA certification.
Training can be a barrier to the diffusion of a new diagnostic. The view of Simon Kimber, project manager of Metro-POCT – a training initiative in the north-west of England – was that the greater the burden involved in organizing training, the less likely it is that a provider will adopt the technology. Typically, training is provided by device manufacturers; Carol Cresswell, POCT Manager for Newcastle Hospitals in England reported that their contract with manufacturers will often include provision for training “cascade trainers”, who can train colleagues within the hospital, as well as periodic top-up training during the term of the contract. However, a risk of relying solely on manufacturers as trainers is that the quality of training may not be adequate and they may omit critical information, such as device limitations and contraindications. Lynda Petley, Manager of the POC Testing Team at Frimley Park Hospital in England, witnessed a trainer employed by a manufacturer informing users that their device was so simple users can never make a mistake. This led staff to believe inappropriately that the system was foolproof. In her view, there is a case for training being independent of manufacturers and, where training is manufacturer provided, service commissioners should be prepared to mitigate the risk by monitoring the delivery of training.
Access to adequate training in primary care was a particular concern highlighted by interviewees. In secondary care, because of proximity to the lab, laboratory staff are more likely to be aware of the introduction of a new diagnostic and are in a position to offer support; in primary care, however, a GP practice will often procure a device without the lab’s knowledge.
Based at Manchester Metropolitan University, Metro-POCT is a two-year proof-of-concept project to improve access to training on POC devices in the northwest of England. Their philosophy is to complement rather than duplicate the training provided by manufacturers and, as one part of their deliverable, they are aiming to work with manufacturers to offer a comprehensive “one-stop” training solution that is simple for users to arrange. Only about 20% of the training that Metro-POCT provides relates to practical device use, the remainder of the curriculum covers device storage (including expiry date checking of consumables and cold chain requirements), sample handling, interpretation of results, clinical waste disposal, health and safety, quality control arrangements, standard operating procedures, and how to integrate use of the device into normal clinical practice. A challenge that project manager Simon Kimber identified as an independent training provider was coping with the variation in device design between different manufacturers.
Some hospitals also offer training to primary care facilities as a commercial offering. Frimley Park Hospital has a team of POC device trainers and offers training and competence testing on request to GP practices. Lynda Petley, from Frimley Park Hospital, viewed offering training in this way as a valuable aspect of a quality-managed POC testing service; not all hospitals will have the resources to provide this service, however, and where it is not mandatory there are problems encouraging GP practices to arrange adequate training, particularly where they need to independently fund it.
A significant proportion of the cost of delivering face-to-face training is the time for a trainer to physically attend a site; both initially, when new staff members join and then for periodic refresher training to support re-certification. One approach, piloted in 2008 by the United States Department of Veterans Affairs (VA), was to deliver training on use of an HIV POCT over the internet to a remote primary care facility using a range of tools, including a webcam.25 Participants in the pilot were satisfied with the approach, which proved to be a practical solution to the challenge of delivering training to the VA’s geographically remote facilities and was more cost–effective than in-person training. The researchers also found that, in the six months after the training was delivered, there was a significant increase in the number of tests being undertaken at the facility. Five years on, Dr Herschel Knapp, one of the researchers behind the original study, reported that online training was still being delivered where appropriate, both for initial training and re-certification, and the benefits reported in the original study appeared to be sustainable. Rather than being the default option, online training has become one of a menu of training methods that the VA uses to meet the needs of either POCT operatives or their trainers. In organizing training, a lesson shared by Dr Knapp was the importance of scheduling training, ensuring that it is organized for a dedicated training time slot when all staff are available or alternatively organized in shifts to ensure that all staff members have the opportunity to attend.
The task of ensuring that all users have received appropriate training can be considerable; Lynda Petley, Manager of the POC Testing Team at Frimley Park Hospital, said her team is responsible for 27 different devices across four hospitals, with training requirements specific to the nature of an individual device. Over 4000 staff are trained annually, with update training and competence checking dependent on the complexity of the device and the frequency of use. In the event of clinical mismanagement of a patient based on a POCT result, the hospital needs a full audit trail to investigate the incident; this could include every aspect of the test, including who undertook the test, when they were trained, who trained them, trainer competence and who maintained the device. In England, each hospital has a different approach to how they manage assuring user competence; in some hospitals, records of training and competence assessment are stored as part of a staff members’ employment record. To support governance, Lynda Petley was keen to see POCTs including secure login functionality to ensure that only centrally authorized users who had proved their competence could access devices.26
As mentioned earlier, overall QC requirements for devices are becoming more burdensome27 and could increasingly serve as a bottleneck in the market for devices requiring extensive QC measures. James Nichols illustrates the extent of the newer requirements.
Traditionally we have bottles of reagents in our chemistry analysers in the laboratory, and we do testing out of those bottles of reagents periodically to make sure that those reagents are still good.28 We do two levels of liquid quality control once a day, and that would tell us if that bottle is still good on our analyser until that bottle runs out, and then we’d run quality control on the next bottle when we opened it up, and periodically until we finished that bottle.29 But now that we have single-use cassettes, we’re running quality control on that cassette. It uses up the whole test and you don’t know that the next cartridge is actually going to behave the same as the cartridge you ran the quality check on. How do you perform quality control on those cartridges?30 You don’t. You quality-control the lot. You insert internal control processes from the manufacturer with each and every test. So this is a different strategy for running quality control that I think is going to challenge the regulatory process in terms of approval. The question becomes, ‘Is this safe and effective when we put it in the hands of general users that don’t have a lot of laboratory experience?’31
The greater quality control also extends to multiplex assays.
Consider the DNA chip that may contain 500 different tests. Do you have to run two levels of quality control on each and every one of those 500 spots on that array? Or, is it sufficient to run a couple of process controls on this card and say that the card is working appropriately? That question is open, and it’s still being debated. But of course there are certain processes for which it is physically impossible to test every aspect of a card, and this is going to come down to risk management.
More and more it will be the laboratory directors determining what is effective in their settings for the way that they are using those test results, and the specific control processes for that device, as factors for how they will manage quality control. The balance between internal engineered control processes on the device and the liquid quality control that the laboratory is analysing, plus the frequency of that quality control, is going to be the responsibility of the lab director more as this risk-based quality control gets implemented.32
Finally, related to the issue of training is the idea that the expertise of clinical staff is only maintained if there is sufficient frequency of use of the device. So while there may be demand for a particular technology, and indeed the use of it could really improve patient care, ongoing training costs may not be justified unless a critical mass of relevant cases exists.
8.3.2. Reconfiguration of pathology services
POC diagnostics are considered to be “disruptive technologies”, or technologies able to have a significant impact on the systems into which they will be introduced. Where the introduction of POC diagnostics is not properly managed, it can leave technicians working in central laboratories concerned about how it will impact on their roles and impact on the quality of patient care, for example through clinicians misinterpreting test results because of insufficient training or out-of-date tests being used due to inadequate quality assurance procedures.33
While some interviewees reported good communication and engagement between laboratory and clinical staff, others felt that engagement was inadequate. Among concerns cited by numerous interviewees were instances of representatives from device manufacturers providing equipment to clinicians directly without the relevant POC testing manager in the laboratory being aware. In the United Kingdom, to prevent this problem from occurring, the industry organization, BIVDA, has issued guidance to its members to “develop and rigorously enforce a policy of involving POCT Managers in the initial stages of marketing POCT products in secondary care, primary care and the community”.34 A particular frustration of one laboratory representative interviewed was that there had been occasions where the laboratory could have offered the same turn-around times as a POC diagnostic, but clinicians had never raised turn-around times for that particular test as a problem before becoming aware of a POCT.
One clinical staff member, who had sought to get laboratory “buy-in” to introducing a POCT across multiple hospitals, felt that a key challenge was securing the trust of laboratory staff, who may take ultimate responsibility for use of the device in their hospital. Clinical staff reported that, in the early days, when there was limited awareness of POC testing, gaining laboratory support could be difficult; now, as long as there is good evidence to support adoption, they didn’t expect many problems. They did, however, report experiencing exceptional cases where they perceived laboratory staff to be blocking the uptake of a diagnostic, regardless of how strong the evidence base to support adoption; they attributed this to “human factors of ego, power and control”, influenced by the attitude of the laboratory director, for example whether the latter was progressive or a traditionalist.
One approach tried at the St. Alexius Medical Center, Bismarck, North Dakota, involves bringing laboratory and nursing personnel together to jointly direct the hospital’s POC testing programme.35 This delivered a range of benefits. By involving nurses in the evaluation of new tests, potential problems were identified before instruments were purchased. For example, there were instances of nursing staff finding devices difficult to use that laboratory staff, with a different type of experience, considered simple. Working jointly also helped facilitate nursing staff’s acceptance of new devices with the view that “nurses were more willing listen to nurses”.
One industry representative interviewed for this study believed that industry has historically taken the approach of marketing POC testing as an alternative to laboratory testing when the more appropriate approach is as part of a unified pathology service with laboratory staff and clinicians working hand-in-hand.
A wide range of resource and organizational considerations must be taken into account in implementing POC testing. These include changes in staff roles and responsibilities, training and competency assurance, putting in place new processes (for example, for ordering and storing test consumables), results handling, quality assurance, ensuring appropriate arrangements to dispose of clinical waste, and equipment storage (for example, if the test is bulky or requires cold chain storage).
While POC testing has the potential to empower clinical staff, it can also be viewed as a burden, an additional duty that is introduced without staffing being increased.36 In a review of the introduction of a POC blood analyser in a rural hospital in New Zealand, respondents reported increased workload. However, views were conflicted: while there was an argument that wards were busier because staff were managing patients who previously would have been transferred, this was balanced by patients being discharged more quickly because staff had direct access to the POC device.37 One interviewee cited personal experience of concerns being raised by a nursing union that the introduction of POCT was overburdening staff. A related point emphasized by another interviewee suggested that, in giving nurses a greater number of duties surrounding diagnosis, POC testing can become an unwanted distraction from proper patient care.
In the United Kingdom, a 2006 independent review found that POC testing was contributing to the fragmentation of pathology services in England, with testing increasingly being undertaken by clinical staff without any reference to pathology practitioners.38 Recommendations made in the associated report included reviewing the role of the pathology workforce, for example pathology staff providing advice on the use of POC diagnostics and taking responsibility for quality assuring decentralized services. Six years on from the report being published, the NHS is continuing to work towards implementing these recommendations.
8.4. Diagnostic and clinical guidelines
Beyond regulatory and reimbursement challenges, clinical awareness of effective diagnostic technology is critical for diffusion of appropriate technologies. Clinical guidelines, or the use of “best-practice tariffs”, which may incorporate appropriate diagnostic use into the accepted care pathway, are potentially important tools for supporting the uptake of new and effective technologies, given the ability of good guidelines to impact clinical behaviour.39 An Audit Commission report on best-practice tariffs, however, found that a detailed knowledge of these tariffs was “not the norm” among clinicians they engaged with for their report, with differing views as to whether educating junior clinicians on them was useful or not, given their decision-making was more likely to be driven by medical evidence than the financial incentives associated with best-practice tariffs. Only one foundation trust had made achievement of best-practice tariffs a component of the medical director’s objectives.40
In addition to guidelines on the appropriate use of diagnostic technology, advice on discontinuation of old/less effective technologies is also seen by many industry developers, and funders, as critical for uptake of new technology: guidelines should look to incorporate direction on this side as well as addressing uptake of new technology. Cost–effectiveness of new platforms is reduced if it is not possible to decommission existing diagnostics, for example due to contractual commitments with suppliers. This may particularly be the case where hospitals have either been loaned, or leased, diagnostic capital equipment at favourable rates by the manufacturer. The placement of equipment within facilities is then linked to contractual agreements to perform a minimum level of tests or volume of consumables such as reagents. This “razor-razorblade” model is frequently employed by manufacturers where the analytical platform is particularly expensive and/or complex, for example as may be the case for many molecular diagnostics. As these agreements tend to be used where the primary platform is of prohibitively high cost for hospitals or labs to purchase outright, the relevance of this issue to POC testing will also likely depend on how expensive the testing device is, particularly relative to the cost of the consumable element. A number of studies have attempted to address the impact of clinical guidelines on medical practice. A recent Cochrane review41 found that the majority of 27 studies evaluating the use of clinical pathway maps showed they had an impact on reducing LOS and hospital costs. There has been little systematic analysis, however, of the impact of guidelines on procurement of diagnostics in the medical field, and this is an area that may warrant further investigation.
Guidelines are one tool that can be used to support bridging the gap between evidence and practice. They are an imperfect tool, however, with evidence of the ability of guidelines to change behaviour considered to be limited.42
In the United States in 2010, 377 children were born with syphilis, a disease that can cause fetal or neonatal death.43 Although the CDC recommends that all women are screened for syphilis during pregnancy,44 this does not always happen in practice. In Florida, where state law requires a minimum of two syphilis tests as part of routine prenatal care and, in some circumstances, a third test at delivery, researchers found that screening guidelines were rarely being followed, with the majority of patients being screened only once and in some cases not at all.45 The reasons for this included poor understanding of guidelines and a lack of awareness that syphilis was a problem, possibly linked to physicians not having encountered cases in practice. A similar picture has been seen with chlamydia screening; despite clear guidance from the CDC, in the United States, less than half of eligible women were screened for chlamydia in 2007.46
In a study involving semi-structured interviews of 20 GPs from Sweden, researchers found variation in perceptions of the link between treatment for UTIs and antibiotic resistance; while some GPs recognized the need to be careful to avoid unnecessary antibiotic prescribing, other views captured by the study were that resistance was, “no problem, I have never seen resistance” or that “the problem is bigger somewhere else”.47 Importantly, only those GPs who did recognize the risk of resistance indicated that they followed all relevant prescribing guidelines.
A variety of reasons have been cited for poor adherence to clinical guidelines; Cabana et al. identified seven key overlapping themes: lack of awareness, lack of familiarity, lack of capacity to comply, disagreement with the guideline’s approach, lack of confidence that the guideline will deliver the relevant outcomes, inertia regarding previous practice as well as external factors, which were divided into three classes: guideline-related, patient-related and environmental.48
Once routine practices are in place, behavioural inertia can make it difficult to bring about change.49 Routine practice can slow the pace of replacing older technology, even when better technologies exist.
In the United Kingdom, a frustration of some industry observers is that there has been no substantive sanction if health care organizations do not follow guidelines from NICE on new technologies that are cost–effective. To address this, the Department of Health has recently committed to introducing a compliance review, which will include publishing an “Innovation Scorecard” to improve transparency of the extent to which local organizations are adopting NICE-approved technologies. While the initiative holds promise, it is still in the process of being implemented.50
8.4.1. Clinical guideline development in the United States
Many guidelines in the United States context are informed, or commissioned, by the AHRQ. At a national level, the AHRQ manages the EPCs programme, awarding five-year contracts to institutions in both Canada and the United States to serve as EPCs. These centres conduct systematic literature reviews on topics of interest to AHRQ, and produce evidence reports or technology assessments. These assessments and reports help inform coverage decisions, as well as guidelines and quality measures for public and private health care payers and providers. The inclusion and exclusion criteria for the reviews are found on most of the EPCs’ web sites and usually frameworks adopt a hierarchy of evidence, where data from RCTs are weighted most heavily.
The AHRQ also hosts the National Guidelines Clearinghouse, a public resource point for evidence-based clinical practice guidelines created or issued by clinical specialty groups or organization. Rather than presenting a definitive guide, the clearinghouse facilitates greater evidence-based comparison between guidelines and identifies areas of conflicting guidance, as well as differences in methodologies. A search for guidelines pertaining to bacterial infections retrieves 170 guides, including guides from United Kingdom-based clinical groups. These guidelines can then be presented in a way that facilitates comparison. For example, the guideline on diagnosis and management of lower UTIs compares and contrasts guidelines from three different clinical groups: the American College of Obstetricians and Gynecologists, the Infectious Diseases Society of America, European Society for Microbiology and Infectious Diseases, and the Society of Obstetricians and Gynaecologists of Canada.
Additionally, within the AHRQ the Effective Health Care Program creates summaries about the risks and benefits of alternative treatments for health conditions, based on comparative effectiveness research (CER). Topics are suggested by the public. Summaries are aimed at consumers, clinicians and policy-makers, yet the disclaimer that the research summaries are not clinical recommendations or guidelines seems to undermine their purpose.
The United States Preventative Services Task Force (USPSTF) is a national body formed of experts in prevention and evidence-based medicine. Guidelines are available online, along with the date of issue and the “status” of the guidelines.51 Guidelines rated A or B are likely to lead to NCDs by federal payers, as well as more likely to achieve coverage with private payers. Screenings for most sexually transmitted diseases are recommended for high-risk populations, along with specific recommendations for molecular testing methods.52 A white paper by UnitedHealth53 shows high volumes of molecular tests for infectious diseases within the Medicaid health maintenance organizations’ beneficiary population. Interviews with UnitedHealthcare representatives confirmed that clear guidelines and coverage contribute to high volumes of these relatively cheap and simple “bread-and-butter” molecular tests being used by providers.
The AMA also hosts guidelines on their web site. For example, in response to the many guidelines for acute respiratory tract infections, the California Medical Association Foundation’s Alliance Working for Antibiotic Resistance Education (AWARE) produced summary guidelines for adults, and paediatrics by synthesizing the available evidence and incorporating the opinions of medical experts and professional organizations. The AMA collaborated with AWARE to host these guidelines on their web site.54
The CDC is a particularly important source of guidelines pertaining to infectious diseases. A report by expert consultants from the CDC explores the use of rapid molecular testing to detect drug-resistant TB and extensively drug-resistant TB (MDR/XDR TB) as a matter for public health concern.55 This report suggests that the CDC should engage with existing laboratory contracts to allow molecular tests to be included for at-risk patients, including those already very ill, or suffering from comorbidities. The management of drug-resistant TB begins with identifying the bacteria reliably, as well as its drug susceptibility. Traditional measures can take up to six weeks, while molecular methods can be as quick as one day. The study also identifies funding challenges associated with rolling out molecular testing, including accurately projecting the volume of tests demanded and associated costs. They suggest that establishing high-volume regional laboratory capacity would improve the viability of this service, but that more research into the costs and benefits is needed. Additionally, they note that there is no evidence of any one test’s superiority and that procurement choices would depend on cost, throughput, turn-around time and validation of the test. This report was sent directly to physician stakeholders, with whom much demand-influence resides, and is accompanied by an analysis of the use of molecular detection of drug resistance, including pros and cons.56
Health care providers are usually governed by the coverage decisions issued by the medical directors of health plans and Medicare contractors. These are nationally determined, locally determined, or determined on a case-by-case basis. Even with a large insurer like BCBS, there may be some coordination of guidelines across regions but each is separately licensed and therefore able to make independent guideline decisions. Clinical specialty groups, advocacy groups and patients can motivate for a specific treatment process to be adopted within their spheres of influence, and many coverage decisions come down to individual decisions about individual patients. These specialty organizations and advocacy groups were mentioned as important stakeholders in determining how and which guidelines are adopted.
Health plans are responsible for defining what medically necessary and reasonable care includes, without curtailing the autonomy of clinical providers. Respondents from health plans and research bodies suggested that this can sometimes be a difficult balancing act and highlighted the variability that exists in how health care is delivered across regions. It was suggested that providers in the medical research “meccas” are often quicker to adopt innovative products and new treatment guidelines, while isolated areas are often left behind, highlighting the need for improved information diffusion and continuing medical education.
8.4.2. Clinical guideline development in the United Kingdom
In the United Kingdom, a number of different organizations are involved in writing and publishing clinical guidelines. The main national-level bodies are NICE in England and Wales, the Scottish Intercollegiate Guidelines Network and the Guidelines and Audit Implementation Network in Northern Ireland. NICE tends to act in a coordination role with relevant clinical bodies in the development of national guidelines, rather than produce them directly themselves given the importance of engaging the relevant clinical stakeholders in the process. In addition to the aforementioned bodies, a number of professional organizations and royal colleges produce guidelines on their respective areas of expertise. NICE has engaged in several pieces of work that include appropriate use of diagnostics largely in the field of cardiology, although little to date in the field of infectious disease. In 2009, NICE established the Diagnostics Assessment Programme (DAP), which is a sister programme to their existing medicines and devices health technology assessment processes. This has been relatively limited in scope thus far. Table 8.1 shows completed assessments and those which are in progress. While the primary output from the DAP is not to write specific clinical guidelines on the use of the evaluated diagnostic technology, the evidence assessment process may be used to contribute to guidelines on diagnostic uptake and use formulated either by NICE other relevant bodies.
Table 8.1
NICE DAP diagnostics guidance to date (published and in progress).
An important tool supporting the appropriate use of diagnostics within the NHS are the “Map of Medicine” pathways that have been developed for a number of indications, with over 1400 local care maps in place across the NHS.58 While the majority of these maps are locally designed, they draw on best-practice clinical advice from professional bodies and NICE clinical guidelines. The assessment of the benefits of introducing these maps in particular settings may encourage evidence sharing across local NHS bodies, and for some indications national care pathway maps have been developed, including for C. difficile and community acquired pneumonia in the field of infectious disease. These process maps set out for the relevant indication at what stage examinations, diagnostics, treatment and referrals should be employed. Therefore an integral part of these maps will be to inform clinicians when it is appropriate to employ a diagnostic to support the patient evaluation and clinical decision-making process. In most cases, however, the maps do not offer specific direction on accessing or conducting the test,59 meaning the guidelines may support the use of some form of diagnostic, but will not necessarily support the use of innovative products per se. A recently announced collaboration between the Royal College of Pathologists and Map of Medicine60 may in time address more specifically the issues around guidance on diagnostic use in the published patient care maps.
In some cases, where there is consensus on best practice, international guidelines may also be used by clinicians. For example, the campaign Surviving Sepsis has actively promoted use of its guidelines on the management of sepsis, and these are used by some hospitals as a benchmark which actual treatment actions are audited against, although not necessarily with any associated rewards or penalties or rewards for (non-) adherence.
While guidelines may support or steer clinical decision-making, in the United Kingdom the NHS executive states that guidelines are not to be used to mandate, authorize or outlaw treatment.61 In some jurisdictions, guidelines hold more legal sway; for example, in France, guidelines published by l’Agence Nationale pour le Développement de l’Evaluation Médicale constitute an enforceable code of conduct for doctors working under the social security system.62
8.5. Prescribing culture
In the absence of rapid POCTs, it takes approximately 36 to 48 hours to undertake a culture and sensitivity analysis using current methods.a, 63 This can leave physicians with no choice but to blindly treat a patient, increasing the risk that antibiotics are prescribed unnecessarily. For example, studies have shown that uncertainty in distinguishing between acute bronchitis and pneumonia has resulted in overuse of antibiotics in primary care.64 Similarly, difficulty differentiating between viral and bacterial causes of conditions such as acute sinusitis65 and otitis media66 has led to inappropriate prescribing.
From the patient perspective, there is significant evidence of misconceptions about the value of antibiotics in treating conditions such as viral infections. In a large population-based telephone survey across seven states in the United States in 1998–1999, over a quarter of respondents erroneously perceived that antibiotics could help cure a common cold.67 In a study from England in 1997 involving 1000 patients with acute lower respiratory tract illness, the majority of patients believed that they had an infection, that antibiotics would be of benefit, and they expected a prescription.68 In this study, no relationship was found between the severity of symptoms and the patient demand for antibiotics.
Where prescribers perceive that patients expect antibiotics, numerous studies have shown that patients are more likely to receive one.69 In an Australian study, patients were 10 times more likely to be prescribed an antibiotic where their GP believed that this was what the patient was expecting.70 In a study by the University of California involving 10 physicians and 306 patients between the ages of 2 and 10, where physicians perceived that a parent expected an antibiotic, antibiotics were prescribed 62% of the time compared to 7% of the time when the physician did not believe an antibiotic was expected.71
This phenomenon may in part be explained by the results of a study in general practice in Wales, which found that despite prescribers being aware that antibiotics may offer limited clinical benefit for sore throats or upper respiratory tract infections, GPs knowingly prioritized the possible immediate benefit to their individual patients above the theoretical longer-term risk to the community of increased resistance.72 In addition to the potential clinical benefits, the act of prescribing may itself help parents with sick children by reassuring them that their concerns have been taken seriously by the prescriber.73
Another factor that has been found to influence prescribing is a prescriber’s level of risk aversion; for example, researchers from Belgium have provided evidence of a link between the degree to which an individual prescriber copes with uncertainty and antibiotic prescribing rates.74 In addition to personal traits, a defensive attitude can be engendered during medical training75 and may be influenced by the structure of the health system. In a multi-country European survey of risk-taking attitudes in 1990, 60% of physicians in Belgium reported that they sought to avoid risks compared to only 24% in The Netherlands, a difference that researchers attributed to greater patient choice of physician in Belgium at that time.76 Where physicians are financially incentivized to ensure patient satisfaction, they may factor in the risk that the patient may switch physician if their expectations aren’t met when making prescribing decisions.77
An approach that has been found to be successful in improving prescribing practice is jointly targeting both patient and prescriber education, for example a community-wide educational intervention in 2002 directed at both health professionals and the public in Tennessee led to a reduction in antibiotic prescribing for children.78 As well as reducing uncertainty in diagnosis, POC diagnostics can also play a role in managing patient expectations. In a multi-country European study on physicians’ and patients’ views on POCTs for lower respiratory tract infection, the most commonly cited advantage of testing was that it could help the physician in managing patient expectations.79 Some of the physicians interviewed also believed that testing could help shift norms and alter patient expectations over time.
A particular concern voiced in the United States market is that socio-cultural factors that encourage defensive prescribing practices have been compounded by the medical-legal environment.80 While there are no definitive national statistics on the volume of medical malpractice claims in the United States,81 it has been estimated that the average physician is affected by an unresolved claim for approximately 11% of their career,82 with the risk of facing a claim differing by specialty. Physicians in low-risk specialties have a 75% chance of facing at least one malpractice claim during their career while the probability of facing a claim in high-risk specialties has been estimated at 99%.83
Using data from the National Practitioner Data Bank, medical malpractice insurer Diederich Healthcare estimated that in 2011 approximately US$ 3.7 billion was paid out to patients.84 Taking into consideration additional costs such as indemnity payments and insurer overheads, legal expenses and the cost of physicians practising defensive medicine, researchers from Harvard University have estimated that the total annual cost of the medical liability systems in the United States is US$ 55.6 billion in 2008 dollars (excluding indirect costs, such as lost clinician work time and the reputational and emotional toll from dealing with cases).85 This equates to 2.4% of total health care costs and was significantly lower than an earlier study by PwC, which used a different methodology and estimated that the medical liability system represented 10% of health care costs.86
Scenarios that have led to malpractice claims include failure to prescribe antibiotics; failure to monitor patients taking antibiotics; antibiotics being prescribed at a suboptimal dose or for too short a period of time; antibiotics being prescribed alone without additional treatments such as surgical drainage and antibiotics being prescribed without reference to diagnostics tests that could have helped identify the more effective class of antibiotic for that particular infection.87
In an online survey in 2012 by the United States health care staffing company Jackson Healthcare, over 1500 physicians were quizzed on their reasons for practising defensively. Reasons cited included “to avoid being named in a potential lawsuit” (78%), because “defensive medicine has become the new standard of care” (61%), because the “patient or family demands that everything humanly possible be done” (59%), “to protect my good name” (48%) and because they believed they were “trained to practice defensively” (19%).88
Although difficult to reliably measure, there are indications to suggest that defensive medicine could be a significant problem; in one 2005 survey of hospital physicians in Pennsylvania, published in the Journal of the American Medical Association, almost all of the physicians surveyed reported practising some form of defensive medicine and 33% of respondents indicated that they often prescribed more medicines than were medically necessary, including antibiotics.89 An independent Gallup poll commissioned by Jackson Healthcare in 2010, found that 73% of respondents practised defensive medicine in the previous year; a practice that Jackson Healthcare has estimated is costing US$ 650–850 billion per year, equivalent to over a quarter of annual health care costs in the United States.90
Where steps have been taken to reduce physicians’ risk exposure, for example limiting the damages that can be claimed for pain and suffering, there is some evidence to suggest that it has been ineffective in reducing defensive practices.91 In Massachusetts, seven hospitals have recently adopted a “Disclosure, Apology, and Offer” policy, which enables physicians to apologize and offer compensation without the admission of apology being used in court92 but, similarly, there is little evidence to suggest that this approach will be effective, with some physicians believing it could increase rather than decrease risk avoidance.93
In recognition of the potential importance of defensive medicine in the fight against antimicrobial resistance, the WHO has identified it as a priority research topic for the future.94
Finally, it should be noted that regional differences in the underlying socioeconomic and cultural attitudes affecting prescribing (and the adaptation of test results) may also be very important but difficult to assess.95 Within Europe itself much variation exists.
8.6. Patient barriers
Patient preferences can also influence the benefits and uptake of POC diagnostics. For example, in a study evaluating the effectiveness of the BioStar Chlamydia OIA POCT, 6.8% of female adolescents tested at a public clinic in Atlanta, were unwilling to wait 20 minutes for the results of the test.96, 97
Footnotes
- a
As of July 2013, these hosts had been announced as the universities of Oxford, Leeds, Newcastle and Imperial College.
- a
It is important to note that in many lower-income countries it often takes up to five days to obtain what amounts to a questionable result.
- Demand-side issues - Ensuring innovation in diagnostics for bacterial infectionDemand-side issues - Ensuring innovation in diagnostics for bacterial infection
Your browsing activity is empty.
Activity recording is turned off.
See more...