Are primary care consultations in Trinidad patient-centered? A cross-sectional study of patients with non-communicable diseases

Background The aim of this study was to measure the patient’s perception of patient centeredness in their consultations for non-communicable diseases (NCDs). We also measured consultation length and patient enablement. Method A cross-sectional study was conducted over 2 months at four primary care clinics at the St. Joseph cluster of the North Central Regional Health Authority (NCRHA) in Trinidad and Tobago. Interviewers timed the consultation and completed post-consultation questionnaires using the Patient Perception of Patient-Centeredness (PPPC) questionnaire and the Patient Enablement Index (PEI). The PPPC is a 14-item (each scored 1–4) Likert-scaled instrument. The total score is averaged and a PPPC score of 4 is the maximum. The PEI measures the ability of the patient to cope with life and their disease. The PEI consists of 6 questions scored 0–2, with a maximum score of 12. Results There were 180 respondents (response rate = 82.5%). Participants were female (75.6%), aged over 65 years (50.6%), married (51.1%), Indo-Trinidadian (52.2%), and Christian (60.6%). Half achieved a primary school education, and 37.2% secondary. The consultation length ranged between 1.32 and 31.22 min. The average, median and mode of the consultation length were 8.5, 7.74 and 10 min, respectively. The average, median and mode of the measures of patient-centeredness were PPPC (3.67, 3.86 and 4) and PEI score (5.93, 6 and 6). The PPPC average was lower in patients with a stroke (p = 0.022), and higher among those with more than 2 consultation interruptions (p = 0.015) and those who knew the doctor very well (p = 0.015). The PEI score was lower in patients with heart disease (p = 0.022). The consultation length was longer in those with tertiary education (p = 0.044) and those with two consultation interruptions (p = 0.032). PPPC Average and PEI Score correlated well (ρ = 0.408, p < 0.001). The consultation length correlated with the PPPC Average (ρ = 0.168,p = 0.025). Conclusion Primary Care consultations in this cluster of health centres in NCRHA in Trinidad were often patient centered. The consultation length, patient-centeredness, measured with the PPPC instrument, and patient enablement scores, measured with the PEI instrument, in consultations for NCDs in Trinidad compare favourably with international reports. Supplementary Information The online version contains supplementary material available at 10.1186/s12875-023-02149-8.

Are primary care consultations in Trinidad patient-centered?A cross-sectional study of patients with non-communicable diseases Reisa R Rahaman 1 , M Shastri Motilal 1 , Raveed Khan 1 and Rohan G Maharaj 1* Patient-centered care (PCC) is defined as "the use of adequate consulting skills in identifying patients' priorities and concerns and the appropriate involvement of patients in making decisions about their care" [1].In this approach the patient's beliefs and characteristics are emphasized instead of the physician or disease-centered approaches where the doctor makes most decisions.Patient-centeredness focuses on three core values: "First, considering patients' needs, wants, perspectives and individual experiences.Next, offering patients opportunities to provide input into and participate in their care; and finally, enhancing partnership and understanding in the patientphysician relationship" [2,3].Patient-centered care is determined by the quality of the communication between doctors and patients.At its core, patient-centeredness envelops therapeutic relationships based on enhanced communication and trust.
The burden of non-communicable diseases (NCDs) is widespread and international efforts are focusing on improving health outcomes attributed to them [14].NCDs cause on average 80% of total deaths in Trinidad and Tobago according to a WHO NCD Country Profile in 2018 [15].Numerous studies have documented that Trinidad and Tobago (T&T) is not doing well with managing NCDs [16][17][18][19][20]. Could patient-centred consultations offer, in part, the solution?The positive impact of patient-centeredness in consultations for chronic disease care has been described [21].For example, a T&T study which used a patient-centered approach with poorly controlled diabetic patients demonstrated statistically significant reductions in HbA1c [13].
There is a paucity of studies on the consultation in the English-speaking Caribbean (ESC).In 1998 a paper reported on the benefit of a Continuing Medical Education (CME) intervention for Trinidadian physicians on patient satisfaction of primary care consultations.Since then, there has been no further published work in the ESC studying the consultation in detail [22].
The evidence is less clear on the value of longer consultation times.The Patient Perception of Patient-Centeredness (PPPC) questionnaire was created to measure the extent to which the patients believed the clinician was patient-centered [23].Another construct, patient enablement index (PEI), has been used as a measure of quality in the primary care setting with good reliability and validity [1,24].Given the burden of NCDs and the importance of the consultation in contributing to success in combatting NCDs, the objective of this study was to determine whether patients receiving care for NCDs perceived their consultation to be patient centered or felt enabled as measured by the PPPC and PEI.The consultation length was also determined.

Study design & location
This was a cross-sectional study at 4 primary care health centres (referred to as #1 -#4 in the analysis and text).They are located, not in the numerical order outlined, at St. Joseph, Tacarigua, Macoya and Arouca.They are all part of the St.Joseph Cluster of health centres in the North Central Regional Health Authority (NCRHA) in Trinidad.The survey was conducted between 3rd October and 22nd November 2018.
A meeting with the nurses and staff was conducted prior to the start of data collection at each health centre.They were informed on the nature of the study and given the pertinent information needed for the completion of this study.A separate meeting was held with clinic physicians to explain the nature of the study.
On clinic days, patients were allocated a number which represents the order in which they will be seen.Consecutive patients were invited to participate in the study.No two patients were included from the same household.
To be included patients had to be at least 18 years, English-speaking and be attending the health centre for an NCD.Patients were excluded from participating in the study if there was any intellectual disability or cognitive impairment, or if they were experiencing an unstable psychiatric disorder.
Any patients with cognitive impairment and intellectual disability are assessed by the doctor and referred to an appropriate clinic, if they were not already enrolled.Such patients are usually accompanied by a relative or caretaker.

Data collection instrument
The data collection instrument contained 4 domains: first, demographics, including age, gender, marital status, religion, ethnicity, education level, household income and employment status; secondly, information pertinent to the consultation including the number of medical problems to discuss, the number of interruptions to the consultation, medical condition(s), the number of medications, doctor familiarity and self-perceived general health status.These questions were derived as part of factors associated with the quality of the consultation from previous studies [1,3,9,22,25].

The patient perception of patient-centeredness (PPPC) survey
The third part of the questionnaire was the PPPC survey.The PPPC average has been proven to correlate with a score covering the above three core values of patientcenteredness, namely 'exploring the disease and illness experience' , 'understanding the whole person' , and 'finding common ground' [23].The questionnaire contains 4 questions pertaining to component 1 (items 1-4), 1 question for component 2 (item 14) and 9 questions for component 3 (items [5][6][7][8][9][10][11][12][13].Each of the 14 questions has a 4-point Likert scale for responses.The sum of the responses divided by 14 is the final score.The final score ranges from 1 to 4; the higher the score, the more patientcentered the consultation.This tool was identified in a systematic review as one of the 2 best instruments to measure the patient's perception of patient-centeredness in the consultation [26].The other instrument was the Consultation Care Measure which had 21 items.The PPPC survey was chosen since it was equal in assessing the patient perception of patient-centeredness but more efficient with 14 items.

The PEI
The PEI score is an outcome measure of patient-centeredness.This means that the more patient-centered a consultation is, "the greater the enablement of the patient to cope with life, understand and cope with their illness, and keep themselves healthy as a result of their healthcare".Enablement was shown to be associated significantly with "interest in effect on life, health promotion, and a positive approach" [1].The PEI questionnaire consists of six questions.Responses are given a score based on a Likert scale ranging from "much better" (2), "better" (1), "same or less" (0) and "not applicable" (0).The maximum score is 12 [24].

Measurement of consultation length
The consultation length has been studied as a measure of patient-centeredness [27,28].The consultation length was measured by one of the interviewers as the patients entered and exited the doctor's room with a stopwatch.

Pilot testing of instrument
Pilot testing was performed at the Arouca Health Centre with 10 patients attending an NCD clinic.After the pilot, the following adjustments were made: No change in the wording of the instrument was required, however the measurement of the length of consultations was carried out by the interviewers instead of the doctor.A copy of the final instrument is available in Additional File 1.

Data collection & analysis
The consultation length was measured by an interviewer using a stopwatch outside the room prior to approaching the patient.The interviewers were volunteer medical doctors who were trained to conduct the consenting process and to administer the instrument.Each patient was approached after they exited the consultation room, the study, and the consent information were explained.Once consent was obtained the questionnaire was administered in a face-to-face interview.
The data collected was analysed using Statistical Package for the Social Sciences (SPSS) Version 23.The main outcomes determined were the mean and SD, median, and mode of the consultation length, PPPC score and PEI score.The frequencies of possible associated factors were tabulated against the ranges for these scores.The associations between patient factors and PEI scores, PPPC Average and consultation length were determined by using the Mann-Whitney U-test and Kruskal-Wallis test.The correlation between continuous variables was tested with the bivariate Spearman's correlation coefficient (ρ).A p-value of less than 0.05 was deemed statistically significant.Post-hoc pair-wise comparisons were used to explore significant associations revealed by Kruskal-Wallis tests.All post-hoc tests conducted were Bonferroni corrected to reduce type 1 error inflation.

Ethical issues
Approvals were attained from the Faculty of Medical Sciences' Ethics Committee (approval # CEC 692/08/18), and the North Central Regional Health Authority, Trinidad and Tobago.All participants provided written informed consent to participate.Seven participants had not completed any formal education (in Table 1); however, these participants were all functionally literate and were able to understand the consent forms, which were read to them by the research assistant.They were all able to sign the consent form.

Results
A total of 180 patients were enrolled.The response rate was 82.5%, ranging from 70 to 100% at the 4 centres.

Consultation parameters
34.4% had 2 medical problems to be discussed, 31.7% used more than 4 medications, and 47.8% gave a general health rating as 'good' .34.4% did not know the doctor at all.Most consultations had no interruptions (80%).See Table 1.

Consultation length
The average, median and modal consultation lengths were 8.5, 7.74 and 10 min respectively, with a range of 1.32-31.22min.The interquartile range was 4.28.The distribution may be seen in Fig. 1.

PPPC average
The PPPC was distributed between 1.79 and 4.00.The inter quartile range was 0.36 and the median was 3.86.The mean was 3.67 with a standard deviation of 0.42.The PPPC was negatively skewed (See Fig. 2).

PEI score
The PEI score ranged between 0 and 12.The interquartile range was 4 and the median was 6.00.The mean was 5.93 and the standard deviation was 3.59.The distribution may be seen in Fig. 3.

Correlations between PPPC score, patient demographics, and consultation parameters
This analysis revealed no statistical differences in the patient perception of patient-centeredness when PPPC scores were assessed by gender, employment status, age, ethnicity, religion, marital status, education, income, number of medical problems to discuss, non-communicable disease, number of medications used and general health rating.The PPPC was determined to be lower in those with stroke (3.10 ± 0.39, 3.07) (Mean ± SD, Median) than those without stroke (3.68 ± 0.42, 3.86) with a p-value of 0.022.The PPPC was significantly higher in health centre 4 (3.76 ± 0.41, 3.93) than health centre 1 (3.57± 0.49, 3.79) with p-value of 0.017.The PPPC was higher in those with more than 2 consultation  The PPPC was significantly higher (p = 0.015) with continuity of care.The PPPC in those patients that knew the doctor very well (3.79 ± 0.35, 3.93) was higher than those who did not know the doctor at all (3.57± 0.47, 3.71).See Table 2.

Predictors of PPPC and PEI scores
Based on other reports from the literature the consultation length was grouped into 0-9 min, 10-15 min, and more than 15 min [29].Analysis suggests that these groupings of consultation times did not have any statistical association with the PEI or PPPC Scores.There were no statistical associations found between the PPPC or PEI and any of the consultation parameters (Table 5).

Spearman's correlations
The correlation coefficient ρ = 0.408 was significant, p < 0.001 showing that PPPC Average correlated positively with PEI Score.The consultation length significantly correlated with the PPPC Average with the correlation coefficient ρ = 0.168, p = 0.025.The PEI Score did not correlate with consultation length significantly with the correlation coefficient ρ = 0.092 (p = 0.220).

Discussion
This study was able to determine the patient perception of patient-centeredness (PPPC), the patient enablement (PEI) and the length of the consultation, among a sample of primary care patients seeking NCD care in North-central Trinidad.From our literature review this was the first time such a study was attempted in the English-speaking Caribbean.
There was some local variation in PPPC scores, for example, the PPPC Average was significantly higher in health centre #4 compared with health centre #1.We postulate that this is because health centre 4 has a lower patient load and therefore could achieve longer consultations (11.39 min vs. 7.85 min respectively).The association between consultation time and PPPC has been shown in previous studies, such as by Howie et al. [1].This reinforces the point that adequate time and continuity of care are both necessary for patient-centeredness to be achieved.
The PPPC Average was lower in patients with a stroke.This is possibly because patients with stroke either need or desire more time and attention.They may also have been experiencing a mood, articulation or cognitive disorder which may lead to lower PPPC scores.
Continuity of care results in greater patient-centeredness [31].In our study, this was seen with a significantly higher PPPC average in those that knew the doctor 'very well' versus those who 'did not know the doctor at all' .A higher PPC score was seen in those with 2 or more interruptions.Consultation times were however longer in those with reported interruptions suggesting that the interruptions did not detract from the overall patientcentered experience.

PEI score-how does T&T compare internationally?
The PEI in our study was determined to be 5.93.In a brief review this was third out of the 9 reports from several   [32], and Sweden with the lowest PEI score of 3.48 (SD = 3.21) [33] and similar to ours, a score from French Canada where the mean PEI score was 5.06 (95% CI: 4.30-5.81)[34].
The differences in PEI in the above countries could be explained by varied primary care needs in diverse populations, cultural variances in inclination to report negative responses, dissimilar understandings of the "enablement" concept, confidence in the doctors in primary care, confidence in the health system and its benefits, continuity of care or genuine variations in the quality of delivered care.
The PEI score was lower in those with heart disease, which may be due to those with heart disease requiring more detailed attention with multiple medical conditions and complex medicine regimes.The variations in PEI among the health centres were not correlated with consultation length or PPPC score.Our study did not find a correlation of PEI with consultation time, which was also found to be a non-significant correlating factor in some previous studies [35,36].

Consultation length-how does T&T compare internationally?
The consultation length average of 8.5 min determined in this study was almost twice the average consultation length (4.6 min for physicians with special training) of a study conducted in Trinidad 20 years prior [22].
Although not recorded in this study, our experience is that in this cluster of health centres about 2/3 of the doctors have completed post-graduate training in family medicine.Communication and consultation skills are integral parts of this training.There are also monthly teachings in the cluster with updates in management.
The consultation lengths in primary care consultations of 67 countries were reviewed in 2017 [29].The range of consultation length was between 48 s in Bangladesh and 22.5 min in Sweden.Only 16 countries had consultation lengths on average more than 15 min.The ranking of T&T is 33 rd among the 67 countries in this international study.
Short consultation lengths are likely to negatively affect patient care and the workload and pressure of the consulting physician.The variances among countries are explained by factors relating to politics and policy, labour force, access, continuity, comprehensiveness, and management.The least effective consultation length is on average 5 min, which at most consists of a greeting and issuing a prescription.These shorter consultation lengths were shown to result in increasing polypharmacy, misuse of antibiotics and inferior communication with patients [37][38][39].
In this study, the consultation length in those patients with University or Tertiary level education was higher than those with Primary school level.This may be due to the greater understanding of their condition by the patient, which in turn may likely result in more questioning of, and discussion with, the doctor.Insufficient time in the consultation is an egregious constraint on the delivery of primary care [40].This is evident by a significant association between shorter consultation and physician burnout.This is postulated to be due to a sense of inadequate personal accomplishment by the doctor in these shorter consultations, which leads to doctors feeling less effective and less capable of managing complex multi-morbid patients [40].This problem needs to be resolved for these patients who require adequate time for effective management in primary care.

Strengths and weaknesses of the study
This study was novel to the Caribbean region and can be used as a baseline and to inform future research in the fundamental aspects of patient centeredness in primary care.
We were able to achieve a good response rate, and an adequate sample size (See Additional file 1), using valid and internationally recognized tools.Recall bias associated with cross-sectional studies was not significant in this case since the consultation they just completed was the one evaluated.
The non-English speaking patients were excluded due to the language barrier however during the study only English-speaking patients attended the clinic.There are very few non-English speaking patients in the sample population.
A weakness of this study is that physicians knew they were being observed.This could have influenced how physicians behaved.However, we see that 70% of    This survey did not measure other factors including physician-related variables that can influence consultation quality, such as fatigue or education.

Recommendations
The training of health care providers in patient-centeredness has been shown in multiple systematic reviews to improve health outcomes [10,[41][42][43].Policy makers need to become aware of these findings.Future postgraduate (PG) education and CME for primary care physicians working in these health centre clusters should focus on the consultation and on quality issues around consultations.Staff should be supported in conducting audits of their practice.There should be further follow up assessment in other health centre clusters of this work with standardization of the PPPC and PEI in this population.

Conclusion
In general, the results suggest that the encounters at primary health care centres in Trinidad are patient-centered, with attention being paid to the patient's illness experience, understanding the whole person, and finding common ground.The patients also in general, had a sense of enablement, i.e., being able to cope with life, understand and cope with their illness, and keep themselves healthy because of their healthcare.The consultation length has roughly doubled compared to a study 20 years ago.

Fig. 3 Fig. 2 Fig. 1
Fig. 3 Histogram of PEI Score and frequency in a population of patients with non-communicable diseases in Trinidad

Table 1
Socio-demographics of the participants and characteristics of their doctor-patient encounter in a population of patients with non-communicable diseases in Trinidad

Table 2
Comparison of PPPC Average and Patient characteristics in a population of patients with non-communicable diseases in Trinidad countries, including Portugal with the highest PEI score of 7.85 (SD = 3.13)

Table 2 (
continued) consultations lasted less than 9 min.If the Hawthorne effect exerted a strong bias, we could expect many more consultations being of longer duration.Weaknesses of the study would include a lack of generalizability of the results to other health centres in Trinidad, the nonresponse for certain questions in the instrument, especially around the PEI questionnaire where many of the responses were 'same' , 'more or less' or 'not applicable' .

Table 5
Consultation Parameters and Patient-centered Measures in a population of patients with non-communicable diseases in Trinidad* Mann-Whitney U-test as indicated*.All others Kruskal-Wallis test *All analyses: p-values > 0.05