Discussion
The findings from this study showed that the interviewees, in the absence of guidelines or consensus, all found their own ways to interpret and handle the iFOBT results—these were perceived as sometimes being helpful and sometimes complicating the matter. A positive iFOBT result could reinforce the decision to refer a patient for further investigation, and negative results were not entirely trusted. The iFOBT results were considered to have less weight than the patient's history. This is in line with the low positive predictive value of this test and an example of probabilistic reasoning.[6,21]
Positive iFOBTs were used as a way to emphasise the need for bowel imaging and were thought to (perhaps) help secondary care prioritise patients. Therefore an iFOBT was sometimes ordered also if there was a history of rectal bleeding. Practical reasons for ordering laboratory tests, among many other factors, were observed in an earlier study.[22]
As CRCs may bleed intermittently, it is reasonable to consider the iFOBT as positive if one of three samples is positive and then to refer for bowel imaging. However, in this study, there were divergent views on what further action should be taken in these cases. Some patients were followed in primary care and never referred to secondary care, which could at least partly explain findings in earlier studies that patients with positive FOBTs were not investigated further.[10–12] There could be several possible reasons for this; for example medical histories considered to be benign, GPs' earlier experiences of many false positive tests, haemorrhoids found at rectoscopies, patients unwilling to be referred for bowel imaging, poor availability of and long waiting times for bowel imaging, as well as the absence of guidelines. After a repeated test that was negative the GP weighed the medical history against the test results and could come to the decision not to refer.
Listening carefully to the patients also means listening to the patient's potential smoke screens and perceived diagnoses, which can mislead the physician.[23] There is thus an underlying conflict in that the physician must at the same time listen to the patient and avoid being misled by the patient. Especially with polysymptomatic patients and those with vague symptoms, it can be a challenge to identify those who need further investigation of suspected CRC.[24]
'Change in bowel habits' was a symptom all mentioned. This expression has a connotation of suspected CRC, but to our knowledge there is no commonly accepted definition of what this means.[25] Many of the participants in this study described that if a patient considered their bowel habits to be changed, then the physicians, too, considered them changed. Thus the patients individually determined the definition.
Continuity of care is generally thought of as a positive factor.[26–28] However, it can also include a risk of delay.[28–30] In this study, continuity and earlier knowledge of the patient was mostly considered helpful, but patients with frequent consultations were thought to be at risk of delay.
Concerns about long waiting times, increased or changed demands from secondary care, and the desire for better communication with secondary care consultants, all of which were described in our study, have also been reported from other countries.[24,31,32] It seems likely that the secondary care consultants would also appreciate better communication.[33]
Deciding whether to refer a patient includes management of uncertainty, and this is an essential part of working in primary care.[34] The interviewees described different ways of dealing with this: ordering laboratory tests, seeking more knowledge, involving the patient in the decision, taking time to reflect, safety netting, referring with safety margins, discussions with colleagues, and asking for a second opinion. With increasing experience, decision-making seemed easier for the GPs in our study, which is in line with the findings of a study in Finland.[35] However, the increased experience did not seem to result in less uncertainty; instead, personal experiences of the difficulties in choosing the right patients to refer appeared to lead to greater cautiousness and humility.
Strengths and Limitations
The physicians in this study were diverse in terms of gender, work experience, and the location of the health care centres; only two physicians who were invited declined to participate, referring to lack of time. The study participants confirmed the contents of their transcribed interviews. We aimed to use well-structured methodology in the analysis in order to add to the credibility of our study.
The number of physicians that were interviewed and the geographic area that was covered was limited. However, the information gleaned from the interviews was plentiful, and we have aimed to describe the context in which the study took place to make transferability judgements possible.
Implications for Clinical Practice and Future Research
Our study results illustrate the difficulties in diagnosing a low-incidence serious disease that presents with common symptoms. The use of the iFOBT as described by the participants illustrates the consequences of using a test in a population for which it has not been evaluated. The iFOBT could potentially be useful as a diagnostic aid in primary care, but more research is needed. Our study also shows the need for evidence-based guidelines and improved communication between GPs and consultants.
BMC Fam Pract. 2015;16(153) © 2015 BioMed Central, Ltd.
© 1999-2006 BioMed Central Ltd
