Skip navigation



Trust: (noun) firm belief that a person or a thing can be relied on; (verb) believe in, rely on.

That’s the dictionary definition of trust, and it perfectly describes the relationship between the NHS and the public over 50 years, for most of the time. That trust is of two sorts. First is trust in the doctors and nurses we see in GP clinics and hospitals, together with an unstated trust in all the unseen support staff in laboratories and offices. Second is trust in the systems of the NHS to deliver the healthcare people need, most of the time.

Despite some hard knocks in recent years, there is still an immense amount of trust in the NHS and in health professionals, and within the NHS between healthcare professionals. Is there anything the literature can tell us about trust?

Literature review

Running a PubMed search with trust as a title word pulls up over 300 references, but most of them are about work done in various hospitals (Trusts) or by charities (Trusts). The trouble is that there is a third dictionary definition, that of a “law arrangement”, and the use of trust in that context dominates the literature.

That doesn’t mean that there is nothing the medical (rather than philosophical or theological) literature can tell us about trust. For instance, there is a meta-analysis [1] about trust in leadership. It tells us that direct leaders (supervisors) are a particularly important referent of trust, and, for those who are interested, “a theoretical framework is offered to provide parsimony to the expansive literature and to clarify the different perspectives on the construct of trust in leadership and its operation”.

Public trust in healthcare systems

The Dutch have taken the issue of public trust in healthcare systems to heart, based on finding no reliable way of measuring it. The idea for the research was suggested by economic research into public trust. So they set out to develop a valid and reliable instrument to measure different dimensions of public trust in health care in the Netherlands [2].

In a first phase, more than 100 people were interviewed to gain insight into the issues they associated with trust. Eight categories of issues that were derived from the interviews were assumed to be possible dimensions of trust. On the basis of these eight categories and the interviews, a questionnaire was developed that was used in the second phase. The questionnaire was sent to 1500 members of a consumer panel; the response was 70 percent.

Analysis revealed that six of the eight possible dimensions were important.

Trust in:

They concluded, not surprisingly that public trust is a multi-dimensional concept, including not only issues that relate to the patient-doctor relationship, but also issues that relate to health care institutions. Their instrument appeared to be reliable and valid. Similar efforts have been directed at measuring patient trust in primary care physicians [3].

Does patient trust matter?

One answer from California would suggest that it does matter, to the patients themselves, their doctors, and to healthcare systems [4].

This was an observational study of visits by 732 patients to offices of 45 physicians (16 GPs, 18 general physicians and 11 cardiologists) in the university and HMO systems in and around Sacramento. Patients and physicians were surveyed before, immediately after, and two weeks after an outpatient visit. Patient trust was measured using a nine-item scale, with items selected from previous patient focus groups and validated trust scales.

Patients were divided into tertiles of low trust, moderate trust and high trust, and outcomes examined according to these tertiles.

The results were interesting. Overall patient trust was high – a mean of 87 when the data were transformed into a scale of 0 to 100, with a range of 14 to 100. Nearly three out of 10 patients rated their trust as perfect. But low trust was associated with lower patient satisfaction and had other negative consequences.

For instance, patients in the lowest tertile of pre-visit trust were much less likely to receive medical information, new medication or a diagnostic test that the requested (Figure 1) or believed they needed. Two weeks after the consultation, this same low trust tertile were less satisfied with the care provided, fewer of them intended to adhere to the advice given by their physicians, and fewer of them had symptom improvement (Figure 2).

Figure 1: Association between pre-visit trust and requested service not being provided (patient report)

Figure 2: Association between pre-visit trust and outcomes at two weeks

Low pre-visit trust was associated more likely to report that they had asked for medical information, though their physicians did not report this association. Both patients and physicians reported higher likelihood of a request of new medication in patients with higher pre-visit trust. There was no other significant association between trust and the request for, or provision of, other services.

So despite there being little difference in service provision between patients, low pre-visit trust meant that patients were less satisfied with their treatment. Their doctors found that the visits were more often more demanding than average in patients with lower pre-visit trust.

Can doctor-patient trust be improved?

There is at least one small randomised trial to attempt to improve patient trust [5]. With only 20 physicians enrolled, even though there was a net improvement in 16 of 19 specific patient-reported physician behaviours in physicians receiving the intervention when compared with those receiving a control intervention, there was, not surprisingly, no statistical significance. Similar results have been seen in a non-randomised study [6].

Trust in general

Perhaps one of the most interesting essays on trust comes from an editorial [7] in JAMA that examined issues of trust between the American Medical Association and JAMA following the departure of George Lundberg as the editor of JAMA. The editorial discusses many topics, and has some interesting comments on academic freedom. For our purposes, though, it has some pertinent words on trust-damaging and trust-building behaviours.

These are summarised in the tables below. Most of us, if pushed, would have come up with similar lists, and while the points in the boxes were directed towards organisations rather than individuals, much the same general principles will apply.

Trust-damaging behaviours

Unwarranted interference
Excessive criticism (especially in the public arena without right of reply)
Coercive or threatening behaviour
Dishonesty or disenginuity
Wilfulness or recklessness

Trust-building behaviours

Mutual recognition of accountability
Shared vision
Explicit strategic objectives
Tactics left unstated
Free and frequent flow of information


Trust is hard won, but easily lost. The increasingly robust examination of scientific and clinical evidence can hardly help individuals at the coal face rather than in pointy-headed academia. There is a much-quoted quote about fundamental issues of trust in scientific papers:

“...only 1% of the articles in medical journals are scientifically sound”.
(Richard Smith quoting David Eddy, BMJ 1991; 303: 798-799.)

We have learned to distrust clinical trials that are not randomised, or not double-blind, or small, or with trivial outcomes. That is why systematic reviews often discard many more studies than they include. Systematic reviews themselves can be flawed, as Alex Jadad and Henry McQuay showed in the field of pain some years ago [8]. Even today there are examples of apparently high-quality research coming to the wrong conclusion because of (apparently) simple errors, or, more often, lack of adequate analytical thought. And neither professionals nor the public are helped by over-hyped media reports of research findings every weekend, alongside advertorials for some snake oil medicine that is just plain daft.


It is not surprising that a distrust rather than trust seems to be the pervading tone. But perhaps "seems" is the operative word. When asked, most healthcare professionals trust their colleagues, most patients trust their doctors and nurses, and even most pointy-headed academics are happy to share control if there is a real problem.

Trust is like quality. We get neither by avoiding hard questions, but when the hard questions are asked and answered, building on what we have gets easier. We know where the rock is, and that we have to avoid the sand.


  1. KT Dirks, DL Ferrin. Trust in leadership: meta-analytic findings and implications for research and practice. Journal of Applied Psychology 2002 87: 611-628.
  2. GF Straten et al. Public trust in Dutch health care. Social Sciences in Medicine 2002 55 :227-234.
  3. MA Hall et al. Measuring patients’ trust in their primary care providers. Medical Care Research Reviews 2002 59: 293-318.
  4. DH Thom et al. Patient trust in the physician: relationship to patient requests. Family Practice 2002 19: 476-483.
  5. DH Thom. Training physicians to increase patient trust. Journal of Evaluation in Clinical Practice 2000 6: 245-253.
  6. DH Thom et al. An intervention to increase patients’ trust in their physicians. Academic Medicine 1999 74: 195-198.
  7. HTO Davies, D Rennie. Independence, governance, and trust. JAMA 1999 281: 2344-2347.
  8. AR Jadad, HJ McQuay. Meta-analysis to evaluate analgesic interventions: a systematic qualitative review of the literature. Journal of Clinical Epidemiology 1996 49:235-243.