Specialist physician knowledge of chronic kidney disease: A comparison of internists and family physicians in West Africa

Abstract Background Postgraduate training is aimed at equipping the trainee with the necessary skills to practise as an expert. Non-nephrology specialist physicians render the bulk of pre-end-stage renal disease care for patients with chronic kidney disease (CKD). We sought to ascertain the knowledge of CKD amongst non-nephrology specialist physicians who serve as trainers and examiners for a training, accrediting and certifying body in postgraduate medicine in West Africa. We also compared the knowledge of family physicians and non-nephrology internists. Methods Self-administered questionnaires were distributed to non-nephrology specialist physicians who serve as examiners for the West African College of Physicians. Results Only 19 (27.5%) of the respondents were aware of the Kidney Disease Outcomes Quality Initiatives guidelines for CKD management. Twenty five (36.2%) of the respondents had adequate knowledge of CKD. There was no significant difference in the proportion of family physicians and non-nephrology internists who had adequate knowledge of CKD (27.3% vs. 40.4% respectively; p = 0.28). Hypertension and diabetes mellitus were identified by all of the physicians as risk factors for CKD. Non-nephrology internists more frequently identified systemic lupus erythematosus as a risk factor for CKD, urinalysis with microscopy as a laboratory test for CKD evaluation, and bone disease as a complication of CKD than family physicians. Conclusion There is a lack of adequate CKD knowledge amongst non-nephrology specialist physicians, since many of them are unaware of the CKD management guidelines. Educational efforts are needed to improve the knowledge of CKD amongst non-nephrology specialist physicians. Guidelines on CKD need to be widely disseminated amongst these physicians.


Introduction Key focus
There is a pandemic of chronic kidney disease (CKD), with the majority of affected individuals being under-diagnosed and under-treated. 1 It has been estimated that CKD affects approximately 10% of the general population. It is usually characterised by progression to end-stage renal disease, and is associated with increased cardiovascular morbidity and mortality. CKD has been shown to be an independent predictor of mortality, with increasing effect as the stages progress. 2 However, timely medical intervention which can slow progression of CKD and prevent end-stage renal disease has been advocated, since the cost of end-stage renal disease programmes is prohibitive. The overall cost of CKD treatment per person per year in 2008 in the United States of America (USA) was estimated to have been between $16 738 and $19 752. 3 The quality of life of patients on dialysis is poor, and the annual cost of haemodialysis in the USA exceeds $60 000 per patient. 4 In the developing world renal replacement therapy is constrained by cost and lack of technological advancements. We previously reviewed the practice and cost of haemodialysis in a teaching hospital in Nigeria, 5 finding that most patients presenting in end-stage renal disease at this hospital -as in many others in Nigeria -are not dialysed because they cannot afford it. Furthermore, dialysis sessions last as long as nine hours due to repeated breakdown of antiquated dialysis machines. The National Kidney Foundation's Kidney Disease Outcomes Quality Initiatives clinical practice guidelines emphasise the need for early detection and management of CKD in order to prevent end-stage renal disease and its consequences. 6

Background
Most individuals with CKD present late to nephrologists -in some instances only when in uraemia. 7 We previously reported on the grim picture in Nigeria, where all patients needing dialysis in a teaching hospital were presenting to the nephrologist for the first time in overt uraemia. 8 This is largely due to late referrals and the dearth of nephrologists worldwide. Taking the USA as an example of the developed world, there were 5500 full-time practising nephrologists as at 2009. 9 The situation is even worse in the developing countries: as at 2009, Nigeria (population over 140 million) had only 103 practising nephrologists. 10 Consequently, preend-stage renal disease care is mostly rendered by nonnephrology specialist physicians. Given the role that nonnephrology specialist physicians play in pre-end-stage renal disease care of CKD patients, it is vital to assess the quality of training that the residency programme imparts to trainee specialist care physicians with regard to CKD, especially in developing countries.

Trends
Suboptimal CKD knowledge exists amongst non-nephrology specialists in the Western world. Published reports indicate that between 35% and 54.7% of non-nephrology specialist physicians in the USA have adequate knowledge of CKD. 11,12,13 Israni and colleagues 11 reported overall CKD knowledge of 35% amongst non-nephrology specialist physicians in the USA. Charles et al. 12 reported that 54% of family physicians in the USA have adequate knowledge of CKD, whilst in a similar study by Agrawal et al., 13 only 54.7% of nonnephrology specialist physicians in the USA had adequate CKD knowledge. Seventy one per cent of physicians in the USA could correctly identify the definition of CKD. 12 Boulware and colleagues, 14 in a study that compared family physicians and non-nephrology internists, reported that 59% and 78% respectively identified the presence of CKD. Lea et al. 15 studied non-nephrology specialist physicians with regard to identification of CKD risk factors, and found that high proportions of respondents identified diabetes and hypertension as major risk factors for CKD; however, only 34.4% identified a family history of CKD as a risk factor, compared to 76.2% in the study by Agrawal and coworkers. 13 In Nigeria few reports exist on the knowledge and practice of physicians regarding CKD. Bosan, 16 working in northern Nigeria, reported poor screening practices for CKD amongst primary care physicians. We recently reported that only 10% of Family Medicine residents attending a workshop in Nigeria had adequate knowledge of CKD and its screening. 17 The majority of respondents in this study did not know the classification and staging of CKD; referrals to nephrologists were done arbitrarily and not based on any particular guidelines.

Objectives
We embarked on this study to assess knowledge of CKD amongst examiners for the West African College of Physicians. We also sought to find out whether there were any differences in knowledge of CKD between family physicians and non-nephrology internists.

Contribution to the field
As there is a dearth of nephrologists, non-nephrology specialist physicians render the bulk of pre-end-stage renal disease care for patients with CKD worldwide. Training and certification of non-nephrology specialist physicians in the West African sub-region is largely carried out by the West African College of Physicians, although two countries (Nigeria and Ghana) have national colleges which also carry out these functions. This study reports the adequacy of CKD knowledge amongst non-nephrology specialist physician trainers and examiners and, by extension, the quality of care that pre-end-stage renal disease patients with CKD in the West African sub-region are likely to receive.

Ethical considerations
The study was approved by the Human Research Ethics Committee of Jos University Teaching Hospital.

Potential benefits and hazards
The subjects were not exposed to any hazards as this was a cross-sectional study that used self-administered questionnaires. Feedback on CKD knowledge was given to the subjects.

Recruitment procedures
Participation was voluntary and consecutive subjects who were willing to participate were recruited.

Informed consent
Informed consent was obtained from all of the participants prior to the study.

Data protection
Data were stored in the Microsoft Excel program, kept secure and only released for analysis when needed. Confidentiality was maintained and the anonymity of responses ensured. Personal identifiers were not collected from the subjects.

Materials
The questionnaire used in this study was patterned after a previously validated questionnaire used to assess knowledge of CKD. 13 Domains assessed in the questionnaire included the definition, staging, risk factors, laboratory evaluation, management, complications and referral of patients with CKD. There were 30 questions in all (a blend of 'best of five' answers and multiple-choice questions of the 'true or false' style), which were used to assess knowledge of CKD. The questionnaire was also designed to obtain information on specialty, gender, practice setting, whether they had a nephrologist in their hospital, whether they saw patients with CKD, and the guidelines they used to manage patients with CKD.

Setting
The subjects for this study were non-nephrology specialist physicians who are examiners for the West African College of Physicians. They were largely drawn from Nigeria, Ghana and Sierra Leone and constituted the bulk of the examiners. They were all holders of the Fellowship of the West African College of Physicians and actively involved in training and certification of residents in the faculties of Internal Medicine and Family Medicine. All of the examiners in the Faculty of Internal Medicine were recruited, except the nephrologists as they were excluded from the study. For every two nonnephrology specialists recruited from the Faculty of Internal Medicine, one specialist was recruited from the Faculty of Family Medicine.

Procedure
The questionnaire was pilot tested amongst doctors in the Department of Internal Medicine at Jos University Teaching Hospital, after which modifications were made to questions and responses as appropriate. The self-administered questionnaires were distributed to the subjects and responses assessed using the National Kidney Foundation's clinical practice guidelines. 6 Adequate knowledge of CKD was defined as answering 21 (or 70%) out of the 30 questions correctly.

Analysis
The results are expressed as proportions for discrete variables and means ± s.d. for continuous variables. The Chi-squared test was used to compare proportions of nonnephrology internists and family physicians with regard to CKD knowledge. The Fisher's exact test was used when cells had less than five observations. The Student's t-test was used to compare the means of the total CKD knowledge scores of non-nephrology internists and family physicians. P-values ≤ 0.05 were considered significant.

Characteristics of the study subjects
A total of 100 questionnaires were distributed, with 69 physicians (47 non-nephrology internists and 22 family physicians) returning completed questionnaires, giving a response rate of 69%. There were 56 (81.2%) men and 13 (18.8%) women. The spread of non-nephrology internists included cardiologists, infectious disease specialists, pulmonologists, endocrinologists, gastro-enterologists and neurologists. Forty-eight (69.6%) of the respondents were practising in university hospitals and 21 (30.4%) in other specialist centres. Fifty-six respondents (81.2%) had nephrologists in their hospitals, whilst 13 (18.8%) did not. Sixty three of the respondents (91.3%) attended to CKD patients in their practices, whilst 6 (8.7%) did not.

Definition and staging of CKD
Only 19 (28.3%) of the respondents were aware of the Kidney Disease Outcome Quality Initiative guidelines for CKD management. Twenty three (34.3%) of the respondents identified the 7th Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) 18 as guidelines for the management of CKD. The remaining 25 respondents (37.3%) were unaware of any guidelines for the management of CKD. Table 1 summarises the findings of our study. Only 26 (38.8%) of the respondents correctly identified CKD, defined as a positive proteinuria test twice in three months. There was no significant difference in the proportion of family physicians and non-nephrology internists who made this identification -8 (36.4%) and 18 (38.3%) respectively; p = 0.87. Only 29 respondents (42%) identified stage 3 CKD as an estimated glomerular filtration rate between 30 ml/min/1.73m 2 and 59 ml/min/1.73m 2 . Eight (36.4%) family physicians and 21 (44.7%) non-nephrology internists correctly identified this staging (p = 0.51).

Risk factors for CKD
All of the respondents identified diabetes mellitus and hypertension as risk factors for CKD. Similar proportions of family physicians and non-nephrology internists identified older age, coronary artery disease, daily use of non-steroidal anti-inflammatory drugs, family history of CKD, male gender and obesity as risk factors for CKD (Table 1). There was a significant difference in the proportions of family physicians and non-nephrology internists that identified systemic lupus erythematosus as a risk factor for CKD (72.2% vs. 97.9% respectively, p = 0.003).

Laboratory evaluation of CKD
Only a few of the respondents (7.2%) would check the serum creatinine level alone as a test of CKD. Fifty eight (84.1%) would check serum creatinine to estimate the glomerular filtration rate. There was no significant difference in the proportion of family physicians and non-nephrology internists in this regard (72.7% vs. 89.4%; p = 0.07). More nonnephrology internists identified urinalysis with microscopic examination as a test for CKD than family physicians (85.1% vs. 59.1%; p = 0.001). Estimation of proteinuria by the semiquantitative (dipstick) method and protein creatinine ratio were identified by similar proportions of family physicians and non-nephrology internists (59.1% vs. 55.3%, p = 0.76; and 36.4% vs. 55.3%, p = 0.14 respectively).

Management of CKD
A total of 61 (88.4%) respondents (81.8% of family physicians and 91.5% of non-nephrology internists; p = 0.24) identified the antiproteinuric effect of angiotensinconverting enzyme inhibitors/angiotensin II receptor blockers independent of blood pressure control. The target

Referral of CKD patients
Twenty six respondents (38.8%) were unsure of the criteria to use when referring a patient with CKD to the nephrologist; only 31 (44.9%) correctly identified the threshold of doing so based on the estimated glomerular filtration rate. The proportions of family physicians and non-nephrology internists here were similar (40.9% vs. 46.8% respectively; p = 0.84).

Total CKD knowledge score
The respondents got 20 ± 5 out of a total of 30 answers on CKD correct. The mean scores for the family physicians and non-nephrology internists were 18.22 ± 5.0 and 20.8 ± 5.0 respectively; p = 0.05, indicating a small difference reaching the border of statistical significance (Figure 1). Only 25 (36.2%) of the respondents had adequate knowledge of CKD. There was no significant difference between the proportion of family physicians and non-nephrology internists with adequate CKD knowledge, as only 6 family physicians (27.3%) and 19 non-nephrology internists (40.4%) answered 21 out of the 30 questions correctly (p = 0.28).

Outline of the results
This study assessed the knowledge of CKD amongst examiners for the West African College of Physicians. We also sought to find out whether there was any difference in knowledge of CKD between family physicians and nonnephrology internists. We found that only a third of the non-nephrology specialist physicians who serve as trainers and examiners for a postgraduate training programme in West Africa had adequate knowledge of CKD. The level of knowledge was similar amongst family physicians and nonnephrology internists, except that the latter more frequently identified systemic lupus erythematosus as a risk factor for CKD, urinalysis with microscopy as a laboratory test for CKD evaluation, and bone disease as a complication of CKD compared to family physicians.
A third of our respondents had adequate knowledge of CKD. This parallels previous finding by Israni and colleagues, 11 who reported overall knowledge of CKD of 35% amongst physicians in the USA. With regard to the various domains of CKD, nearly 40% of our respondents correctly defined CKD. This is comparable to the 54% reported by Charles et al. 12 amongst family physicians and 54.7% by Agrawal et al. 13 amongst non-nephrology specialist physicians in the USA. However, the performance of non-nephrology internists in our study with regard to definition of CKD is a far cry from that reported by Charles and coworkers 12 amongst non-nephrology internists (38.3% vs. 71% respectively). Our findings are also much lower than the 59% and 78% performance of the family physicians and non-nephrology internists in identification of the presence of CKD and its severity as reported by Boulware and colleagues. 14 The proportions of our respondents who demonstrated adequate knowledge in the domain of risk factors for CKD are similar to that reported by Agrawal et al. 13 amongst internal medicine residents in the USA. The study by Lea and colleagues 15 reported similarly high proportions of physicians identifying diabetes and hypertension as risk factors for FP Total Score NNI Total_Score CKD knowledge score CKD. However, only 34.4% of their respondents identified family history of CKD as a risk factor, compared to 69% in our study and 76.2% in that by Agrawal et al. 13 Comparable performances were recorded in our study in the domains of laboratory evaluation of CKD, management of CKD and complications of CKD, with few notable differences. The proportion of physicians in our study who identified urinary protein creatinine ratio as a laboratory test to evaluate CKD was lower than in the study by Agrawal et al. 13 (49.3% vs. 76.2% respectively). Likewise, marked differences also existed between our findings and theirs with regard to proportion of physicians identifying target blood pressure in management of non-proteinuric CKD (50% vs. 89.1%) and coronary artery disease as a complication of CKD (29% vs. 53.7%).
Proteinuria has been shown to be a risk factor for CKD progression, and its amelioration shown to retard progression of CKD. 19,20,21,22 Over 88% of our respondents identified the antiproteinuric effect of angiotensin-converting enzyme inhibitors/angiotensin II receptor blockers. This is similar to the rates reported previously by Israni and co-workers 11 and Agrawal et al. 13 Use of angiotensin-converting enzyme inhibitors/angiotensin II receptor blockers forms the cornerstone of retarding progression of CKD. This is a management strategy that can be employed at all levels of care to effectively reduce progression of CKD.

Practical implications
This study demonstrates the lack of CKD knowledge amongst non-nephrology examiners for a certifying postgraduate college in the West African sub-region. The results of our study have several implications for residency training and the quality of specialists produced by the College, as well as for clinical practice and pre-end-stage renal disease care of patients with CKD. The trainers and examiners do not have adequate knowledge of CKD. This is likely to impact on both undergraduate and postgraduate students in Medicine, the result being a similar lack of adequate CKD knowledge. This in turn is likely to affect the quality of pre-end-stage renal disease care for patients with CKD, since this is mostly offered by non-nephrology specialist physicians. Simple but effective measures like restriction of dietary salt intake, weight loss in the obese and cessation of cigarette smoking utilised in management of CKD were missed by nearly 20% -50% of these physicians.

Limitations of the study
We encountered some limitations that would restrict the generalisability of our findings. We studied only examiners for the West African College of Physicians, and are unaware of the knowledge of the examiners for the Ghana College of Physicians and Surgeons and National Postgraduate Medical College of Nigeria, which also contribute to postgraduate education in Ghana and Nigeria. Whilst attempting to cover all of the domains of CKD, we may have less accurately assessed the depth of knowledge, as the questionnaire utilised closed-ended questions. In addition, the definition of adequate knowledge was arbitrarily set as scoring 70% and above. Finally, the responses of the physicians cannot be readily associated with actual practice, as the knowledgepractice gap is a common limitation of questionnaire-based surveys. Despite these limitations, the strength of our report is that the physicians studied were trainers and examiners at all levels of postgraduate training in the sub-region.

Recommendations
Educational efforts are needed to improve the CKD knowledge of non-nephrology specialist care physicians. Guidelines on CKD need to be disseminated widely amongst these physicians. Practical steps like automated reporting of estimated glomerular filtration rate should be embarked upon by laboratories in the developing world, as this may facilitate early CKD recognition and consequently appropriate referrals.

Conclusion
This cross-sectional study assessed knowledge of CKD amongst non-nephrology specialist care physicians that serve as examiners for the West African College of Physicians. The knowledge of CKD of these physicians was inadequate, as many of them were unaware of the CKD management guidelines. There was no significant difference in the proportion of family physicians and non-nephrology internists with adequate CKD knowledge.