Characteristics of included studies
We included seven studies [
16‐
22] with a total of around 200 participants (see Table
1). They took place in Europe (
n = 6) and Australia (
n = 1). One was published in 1997, and the remainder were published between 2007 and 2011. All used semi-structured interviews to gather data; two also used focus groups. The 1997 study referred to blood POCTs using samples obtained by venipuncture and analysed onsite in the health centre [
19]; the others referred to finger-prick blood tests. All examined attitudes of general practitioners (GPs); two also examined attitudes of nurses.
Table 1
Characteristics of included studies
| Wales (United Kingdom) | Semi-structured qualitative interviews | A test to distinguish bacterial from viral infections using a finger-prick blood test | No experience – participants discussed their perspectives on possible introduction of the POCT | 40 | GPs |
| The Netherlands | Semi-structured qualitative interviews | C-reactive protein POCT for lower respiratory tract infection and other common infections | All participants had been using the POCT for nearly 3 years at the time of interview as part of a randomized trial | 20 | GPs |
| The Netherlands | Semi-structured qualitative interviews | C-reactive protein POCT to differentiate serious from self-limiting lower respiratory tract infection | 10 participants had used the POCT for at least two years at the time of interview as part of a randomized trial; 10 participants had no experience | 20 | GPs |
| United Kingdom | Semi-structured interviews and a focus group | A range of POCTs including haematology (full blood count, platelets); chemical pathology (sodium, potassium, urea, creatine); glucose, cholesterol; bilirubin, alkaline phosphatase, aspartate transaminase; creatine kinase | Participants worked in a health centre where POCTs were piloted; a nurse took blood samples using venipuncture, they were analysed onsite, and the results were made available to the GP at the end of surgery or immediately if requested | Unknown | GPs |
| Australia | Group discussions + individual interviews | INR (international normalised ratio) fingerstick test for monitoring patients on warfarin | No experience (this is not stated explicitly but is assumed) | 33 participants in total; unknown how many were GPs and nurses | Hospital pharmacists, specialists, nurses, GPs. We included only the attitudes of GPs and nurses in the review (nurses treated patients in their homes as well as in hospital) |
| United Kingdom | Semi-structured qualitative interviews | HbA1c (glycated haemoglobin) finger-prick test for patients with type 2 diabetes | Participants took part in a pragmatic, open, randomized controlled trial, where they gave some patients usual care and others POCTs for 1 year | 11 | GPs, practice nurses |
| Belgium, Hungary, Spain, Wales, Poland, Italy, England, Norway, The Netherlands | Semi-structured qualitative interviews | C-reactive protein POCT to aid management of acute cough/lower respiratory tract infection | Participants from Norway routinely used the POCT; participants from other countries had no experience | 80 | Primary care clinicians |
Two studies used data obtained from interviews with GPs participating in the same randomised trial [
17,
18]. Each had 20 participants, 10 of whom overlapped between the two studies. Since both had 10 different participants, and the focus of analyses were different, we included both in our synthesis.
The type of test included in each study is shown in Table
1. Four studies examined attitudes towards C-reactive protein (CRP) POCTs or hypothetical tests which could similarly distinguish between viral and bacterial infections [
16‐
18,
22]: we refer to these as diagnostic. Two examined POCTs for monitoring chronic illness (patients with diabetes [
21] and those taking warfarin [
20]): we refer to these as monitoring. One examined attitudes towards a range of POCTs [
19]. We looked for similarities and differences in attitudes towards diagnostic and monitoring POCTs.
Studies varied according to whether participants had experience using POCTs, were being asked about a test of which they had no experience, or contained a combination of those with and without experience (Table
1). Three studies including participants with experience were conducted in the context of a randomised trial in which a test was introduced as an intervention to all [
17,
21] or some [
18] participants, and one included GPs from a health centre where POCTs were being piloted [
19]. Another included GPs from Norway, where CRP POCTs are routinely used, and from eight other European countries where they are not [
22]. We looked at similarities and differences in attitudes between clinicians with different levels of experience.
Five of the included studies were of good quality (see Table
2) [
16‐
18,
21,
22]. Study samples, data collection and analyses were appropriate and they were clearly described. Another study lacked some details about the sample (for example how many of the participants were GPs and nurses, and whether they had any experience at all of using POCTs), and the sample size was small (only one focus group for each group of clinicians); but the methods of data collection and analysis were appropriate [
20]. We considered these studies to be relatively equally rigorous and trustworthy and treated them equally in the synthesis. One other study was poorly described [
19]: it lacked details including the number of participants; the design, duration and timing of the interviews and focus group; how the data were analysed and by whom; and whether the results include verbatim quotes or not. Therefore it is not possible to assess how rigorous and trustworthy the study is. Therefore, this study did not contribute as much to the synthesis. Additionally of note when interpreting the results of this study, the nature of the POCT in this study differed from the others in our synthesis because it used blood obtained by venipuncture rather than finger-prick blood samples.
Table 2
Quality appraisal of included studies
| Yes | Yes | Yes | Yes | No | Yes |
| Yes | Yes | Yes | Yes | No | Yes |
| Yes | Yes | Yes | Yes | No | Yes |
| Unclear | Unclear | Unclear | Unclear | No | No |
| Unclear | Yes | Yes | Unclear | No | Yes |
| Yes | Yes | Yes | Yes | No | Yes |
| Yes | Yes | Yes | Yes | No | Yes |
Amongst all studies there was a lack of discussion about the impact of researchers’ characteristics and perspectives, and their relationships with participants. This absence of reflexivity limits our ability to assess the influence of the researchers on the data and interpretations.
Through the synthesis we identified descriptive themes regarding positive and negative attitudes of primary care clinicians towards blood POCTs. From these, we developed three analytical themes which are discussed below. Within each of these themes there are issues which may act as facilitators and barriers to widespread adoption of POCTs in primary care: Table
3 summarises the barriers and facilitators to POCT use within each theme.
1)
Impact of POCTs on decision-making, diagnosis and treatment
Table 3
Summary of how primary care clinicians’ attitudes towards blood POCTs may act as facilitators and barriers to their adoption in primary care
Impact of POCTs on decision-making, diagnosis and treatment | Increased diagnostic certainty | Concerns about accuracy |
More effective targeting of treatment (e.g. antibiotics) | Might not be helpful or alter consultations |
| Possible misleading results |
Impact of POCTs on clinical practice more broadly | Fewer re-consultations / phone calls for the same or future episodes of illness | Over-reliance, undermining of clinical expertise |
Enhanced confidence and job satisfaction | Cost, equipment maintenance, time |
Avoidance of missing or delayed results, and loss of patients to follow-up | Usefulness limited to certain situations and patients |
Impact of POCTs on patient-clinician relationship and perceived patient experience | Enhanced communication through discussing immediate results | Possible patient dislike of testing |
Increased patient education and self-management of chronic conditions | Patient anxiety resulting from intermediate results |
Shared decisions with patients (e.g. antibiotic prescription) | |
Greater reassurance and satisfaction for patients | |
Patient confidence in clinicians’ decisions | |
Many attitudes were related to how POCTs might enhance immediate diagnosis and treatment. Diagnostic POCTs were viewed as helpful for improving diagnostic certainty and confidence in clinical decisions; [
16‐
18,
22] particularly for ruling out serious infections [
17].
POCTs were perceived to enable more effective targeting of treatment. Particularly, tests which could distinguish viral from bacterial infections were considered helpful and could aid decision-making regarding antibiotic prescription: [
16‐
18,
22].
“It also helps you to be a bit more careful in prescribing antibiotics, that’s true. It makes you more aware that you may be using them too often” (GP [
17]). This was a belief of GPs who had both used [
17,
22] and not used diagnostic POCTs [
16,
18,
22].
A primary concern was the analytical accuracy of POCTs [
16,
19,
20,
22]:
“the results they give are not accurate enough” (Primary Care Clinician [
22]), which might lead clinicians to miss serious infections [
22]. Clinicians did not feel ‘convinced’ or confident about their performance [
16,
20]:
“we’ve had no research presented to us” (Nurse [
20]). In one study GPs raised concerns that they would be liable medico-legally for any problems arising from inaccurate results [
19] (note this study was poorly described and it is not possible to assess its rigour).
Although POCTs were perceived on the whole to enhance patient care (if tests were accurate), exceptions were noted. A small number of individuals believed that it was not important or always helpful to distinguish bacterial from viral infections, [
16] that monitoring POCTs did not influence the outcome of a consultation, [
21] or questioned the added diagnostic value [
22]. Diagnostic POCTs would not be helpful when serious complications arise from viral illnesses [
16]; and misleading results due to CRP not being raised in the early stages of illness, or due to false results, could lead to inappropriate diagnosis and treatment:
“I see the disadvantage that a mistake or false results can come out as a result. So for instance there is a positive result…. But a different and hidden problem can be the cause” (Primary Care Clinician [
22]). Usefulness of monitoring POCTs performed by nurses varied
“according to the nurse’s level of responsibility for making management changes and the availability of a doctor during nurse-led clinics” (authors [
21]). GPs in one study actually felt waiting for results from laboratory testing was advantageous because it gave them time to
“defer decision-making while awaiting results, thereby ‘allowing nature to take its course’” (authors [
19]) (note that this study was poorly described and trustworthiness of findings cannot be assessed).
2)
Impact of POCTs on clinical practice more broadly
Further to the direct impact on diagnosis and treatment, POCTs were thought to have a wider-reaching impact. The immediacy of diagnostic POCT results could reduce re-consultations or phone calls regarding the same episode of acute illness [
17,
22]. Some GPs believed that consultations for future illnesses may also be reduced:
“If you don’t treat a patient with antibiotics [after CRP testing] and the complaints resolve spontaneously, I think that patients will tend to wait and see and not consult the doctor again for the next similar illness episode. So what we hope is that this management including CRP will lead to fewer consultations or repeat consultations for new infections” (GP [
17]). Advantages of monitoring POCTs in terms of future consultations were
“avoidance of missing or delayed results and occasional loss of patients to follow-up” (authors [
21]).
Immediacy of results could enhance clinicians’ confidence and job satisfaction when using monitoring POCTs:
“My confidence has actually grown in discussing the result with them… I feel it’s sort of added and rounded off the consultation” (Nurse [
21]).
There were some concerns that clinical practice could be negatively affected. Clinicians worried about potential over-reliance on diagnostic POCTs, [
17,
22] undermining of clinical expertise, and over-testing:
“Perhaps it’s being used a bit too often. I think you need to be careful about that” (GP [
17]);
“The disadvantage is that doctors may rely more on test results than on clinical judgement” (Primary Care Clinician [
22]);
“There’s a risk that you let the test determine your management. In the end, what matters is the person who’s sitting there and what you hear and what you find on physical examination” (GP [
17]).
Clinicians also expressed concerns that POCTs could only be used intermittently and in certain situations and patients [
16,
18,
20]:
“for example, in situations where they were unsure of the aetiological cause on the basis of the clinical presentation, or in a situation of deadlock with a patient who definitely wanted antibiotics” (authors [
16]).
Concerns regarding feasibility included cost, [
16,
19,
21,
22] maintenance of equipment, [
16] quality control, [
19] time [
16,
17,
22] and organisational issues (for example interference in nurse activities) [
17]. More positively, POCT devices were described as user-friendly, [
17,
21] and in some cases as having
“very little influence on their [GPs’] workload” (authors [
17]).
3)
Impact of POCTs on the patient-clinician relationship and perceived patient experience
Participants felt that being able to discuss results of monitoring POCTs with patients immediately was beneficial for patient-clinician communication, and determining the most appropriate treatment plan [
21]:
“you can instigate changes in treatment there and then and discuss it with the patient” (Nurse [
21]). POCTs could therefore enhance patient education and self-management of chronic conditions [
20,
21]:
“It’d be great for patient advocacy and empowering them to take some responsibility for their own health care” (Nurse [
20]).
Regarding diagnostic POCTs, it was believed that patients would be convinced, reassured and more satisfied in their GP’s decisions if POCTs had been used, compared to if they had received no test [
16‐
18,
22]:
“then you can justify what you are saying to the patient. Because nowadays, patients want the evidence as well” (GP [
16]). In particular, a test result confirming a GP’s decision not to prescribe antibiotics would help them to “
sell” this decision to patients [
16] and manage patient expectations for antibiotics, [
22] leading to shared decisions with patients [
18]. This was perceived by GPs to help preserve a trusting doctor-patient relationship [
17]. GPs with different levels of experience of using diagnostic POCTs had similar perceptions that they would help to reassure patients and lead to more effective targeted treatment without alienating or upsetting patients [
16,
18,
22]. GPs in one described that the POCT service
“boosted the practice’s image”[
19] (note that this study is poorly described and rigour cannot be assessed).
Although it was widely believed that patients would like to have POCTs available, concerns that patients may not like testing were mentioned by a minority of participants, [
16,
17,
22] with children mentioned in particular [
16]. Furthermore, some GPs were worried about difficulty interpreting and explaining diagnostic test results, [
22] particularly intermediate results [
17] which could increase uncertainty in patients:
“the patient may think that their blood was not entirely OK, so that may make them insecure and worried” (GP [
17]). With regards to interpreting test results,
“a solid training session was highly valued” (authors [
17]).