Ideas, concerns, expectations, and effects on life (ICEE) in GP consultations: an observational study using video-recorded UK consultations

Summary

In most consultations, patients raised at least one ICEE component, most commonly ideas about the cause of their symptoms. Most ICEE events occurred in the 'gathering information’ stage of the consultation. GPs initiated ICEE dialogue in less than half of the consultations, asking mostly about patient ideas and concerns but rarely about expectations. The larger the number of problems discussed in a consultation, the more likely it was for more ICEE components to be raised. Problems that were acute, assessed by older GPs, or assessed earlier in the consultation were associated with higher ICEE scores. There was weak evidence that problems raised by older patients or patients from more deprived backgrounds had lower ICEE scores. Mentions of patients’ ideas and patients’ concerns were associated, respectively, with increased and decreased post-consultation satisfaction.

Strengths and limitations

The screening for ICEE events was robust, with a low threshold for a second review of anything that could be an ICEE event. Using recordings and transcripts is advantageous over direct field observation¹⁴ as it allows multiple reviews of consultation content by multiple coders.

While the study demonstrated associations between patient satisfaction and the presence of ideas and concerns, these results should be interpreted with caution. First, all apart from one patient were satisfied or very satisfied with their consultation and one can question the relevance of comparing these two ratings. However, an interview study of patients reporting these two post-consultation scores has shown a meaningful difference akin to 'acceptable' (satisfied) and 'outstanding' (very satisfied).²² Given previous research²³ has reported similar skewed post-GP-consultation satisfaction distributions (77% ‘very satisfied’), it was felt the dichotomous analysis of ‘very satisfied’ or not was appropriate for this distribution. Second, neither association was detected when only GP-initiated ICEE components were assessed, so these associations may have been related to how ICEE were subsequently dealt with by the clinician. Third, the models were limited by the sample size available, and the development of multiple exploratory models increases the probability of producing a spuriously significant result. Most attention should be paid to the overall ICEE models rather than patient- or GP-initiated specific results that could be impacted by one another (for example, if a patient has already vocalised their ideas there would be no need for GPs to directly ask this question). Finally, patient satisfaction is known to be affected by factors both within and outside the consultation such as the helpfulness of reception staff.²⁴ While adjustments for covariates were made, there may have been additional confounding factors, such as patients’ health literacy,²⁵ and clinicians' ‘positive approach’ and display of empathy, which have been linked to increased satisfaction.²⁶

The analysis was limited to consultations that contained at least one new problem from 23 GPs in the West of England and therefore is only reflective of a subset of GP–patient interactions. An association was not found between ICEE scores and if patients were consulting their usual GP. However, analysis of these data were limited by almost one-third of patients reporting either not having a regular GP or not knowing who their regular GP was. A larger sample size is required to draw more firm conclusions on the relationship between ICEE and other patient or clinician factors.

Comparison with existing literature

Observational studies of ICEE have been reported outside of the UK. Matthys et al ¹⁴ reported on ICE in 613 GP–patient consultations in 36 Belgian practices observed in 2005. They found that in only 20% of consultations were all three ICE components expressed, compared with 34% in the current study. Unlike Matthys et al, the present study did not find an association between the presence of concerns or expectations and fewer prescriptions.

In 2006, Towle et al ¹³ examined audio-recorded consultations involving 198 patients and six family clinicians in Canada. Most ICE events occurred in the ‘problem identification’ rather than ‘problem management’ phase, consistent with the present study. Although Towle found clinicians were competent at eliciting ICE, ‘rarely’ were all three components explicitly addressed. Clinicians reported they made a conscious decision whether to ask these questions citing a lack of time, concerns about repetition, discomfort dealing with patients’ emotional state, and that it could be equivalent to opening ‘Pandora’s box’.

Freilich et al ²⁷ used questionnaires in 2015 to compare perceptions of whether ICE emerged during consultations with GPs, district nurses, and physiotherapists in Sweden. Patient and GP assessments of ICE components were roughly concordant: ideas (76% patient versus 73% GP), concerns (23% versus 17%), expectations (31% versus 40%). These percentages are lower than the present study's findings for concerns (55%) and expectations (51%), but similar for ideas (79%).

Implications for research and practice

Future studies could use patient post-consultation questionnaires to explicitly ask whether patients thought they had discussed their ICEE and whether these were adequately assessed. Likewise, a clinician post-consultation questionnaire could be used to evaluate researchers’ and clinicians’ concordance with respect to classifying ICEE.

Post-COVID-19 pandemic, alternatives to face-to-face consultation in primary care have become more common.²⁸ This study provides a benchmark of ICEE frequencies in face-to-face consultations that could be compared with other consultation modalities.

Further research on how ICEE components are initiated and responded to, rather than just their presence, may provide deeper insights into their impact on patient outcomes. Further analysis to understand why patients in the study who mentioned their concerns were less satisfied would be helpful when training new GPs. One possible explanation is that patients voicing concerns were more anxious pre-consultation, and less easy to reassure. Another explanation could be dissatisfaction with how their concerns were subsequently addressed. The use of concise pre-consultation forms has shown promising initial results in identifying patients' concerns, enabling clinicians to proactively address them and improve satisfaction.²⁹

The finding that problems assessed later in the consultation had fewer ICEE components raised aligns with previous research demonstrating additional problems had less safety-netting advice³⁰ and only increased consultation duration by 2 minutes per extra problem,³¹ adding to the evidence multi-problem consultations may be higher risk or less satisfactory for patients and GPs.

Finally, the finding that older GPs discussed more ICEE components per problem than their younger counterparts, contrasts with a previous study that showed that younger GPs provided more ‘safety-netting advice’.¹⁸ This warrants further investigation into how routine practice is changing with time and any impacts this will have on the quality and safety of consultations.