Physicians make different decisions from nephrologists at serum creatinine 2.0 mg/dl
- First Online:
- Cite this article as:
- Tamba, K., Kusano, E., Tabei, K. et al. Clin Exp Nephrol (2009) 13: 447. doi:10.1007/s10157-009-0176-4
- 33 Views
It is very important, but not clear, how physicians differ from nephrologists in treatment of renal insufficiency.
To demonstrate differences in decision-making in treatment of renal insufficiency between physicians and nephrologists.
Design of study
All physicians were graduates from one medical school and certified by the Japanese Society of Internal Medicine. Nephrologists were certified by the Society and the Japanese Society of Nephrology.
Questionnaires were sent to 1,395 physicians and 385 nephrologists, including audit of serum creatinine concentration that would indicate referral to nephrologist, audit of continuation of angiotensin converting enzyme inhibitor (ACEI) for a case of renal insufficiency and mild hyperkalemia due to ACEI. Outputs were proportion that selected “serum creatinine 177 μmol/l (2.0 mg/dl) and over” as a referral point to the nephrologist, and proportion that chose “suspend ACEI” for a case of renal insufficiency and mild hyperkalemia due to ACEI.
Six hundred and fourteen physicians replied (44%), and 111 certified in internal medicine were extracted from them. One hundred and eighty-six certified nephrologists replied (47%), and 114 certified in internal medicine were extracted. The proportion that chose “177 μmol/l” as a referral point to the nephrologist was 20% for physicians and 61% for nephrologists (P < 0.0001). An additional 17% of nephrologists recommended creatinine concentration below 177 μmol/l, whereas no such opinion was found among physicians. The proportion that chose “suspend ACEI” was 45% for physicians and 16% for nephrologists (P < 0.0001).
There is significant difference between decisions made by physicians and nephrologists regarding treatment for patients with serum creatinine concentration of 177 μmol/l.
KeywordsGeneral physicianNephrologistSerum creatinineChronic kidney failure
The number of patients who undergo renal replacement therapy is increasing at the rate of approximately 10,000 persons per year in Japan. The number of chronically dialyzed patients was about 275,242 at the end of 2007 . This has become a serious medico-economical problem because the total annual costs for dialysis therapy exceeded US $14 billion (¥1.3 trillion) in 2004.
Recently, early referral to nephrologist was recommended in several papers [2–4]. The guidelines of the National Kidney Foundation recommended that glomerular filtration rate (GFR) below 30 ml/min/1.73 m2 should be referred to nephrologist . However, it is still unclear when is the best time.
It is evident that proper use of ACEI, and strict control of blood pressure and blood glucose, can delay or prevent progression of renal dysfunction [6–8]. However, most Japanese physicians do not seem to be competent at treating renal diseases. They have less chance and experience of seeing renal patients than do nephrologists. Moreover, progression of chronic kidney diseases is very slow. It takes years for the severity of the disease to manifest itself, at which point it is already too late. This means that most physicians may overlook the importance of maintaining kidney function. If physicians referred patients to nephrologists at the appropriate stage of renal dysfunction, the number of Japanese end-stage renal disease (ESRD) patients would not increase at such an alarming rate.
We surveyed the differences in decisions made in treatment of renal insufficiency between physicians and nephrologists in Japan, and found that there is a wide gap between them. We draw attention to the possibility of a similar rift between physicians and nephrologists in other countries.
This study is based on the results of a posted, anonymous, self-administered survey. We sent the survey in December 2003 and January 2004. Background data about physicians were collected on gender, career, specialty, workplace, final training in nephrology, numbers of daily renal patients, renal tests used for outpatients and access to nephrologist and dialysis machine. The survey included multiple-choice and/or multiple-select questions about renal diseases and specific cases.
Questionnaires were sent to 1,395 physicians and 385 nephrologists authorized by the Japanese Society for Nephrology. The nephrologists were extracted by random-number selection from a list of certified nephrologists on the homepage of the Japanese Society for Nephrology. Physicians were extracted from the address book of the graduates of Jichi Medical School. Jichi Medical School was established in 1972 to ensure and improve the level of medical services in remote areas where medical resources are scarce. We extracted those certified by the Japanese Society for Internal Medicine from both nephrologists and physicians. We excluded nephrologists who had been working for more than 27 years in order to match years of work in the two groups, since there were no graduates of Jichi Medical School who had worked for more than 27 years at the time of this survey. We also excluded the physicians who subspecialize in nephrology and those who do not currently see renal patients.
In this study, we analyzed data obtained for the following two questions about mild renal insufficiency, shown below.
Q1. Which is your most preferable concentration of serum creatinine or condition for referral to the nephrologist?
177 μmol/l (2.0 mg/dl)
354 μmol/l (4.0 mg/dl)
530 μmol/l (6.0 mg/dl)
707 μmol/l (8.0 mg/dl)
884 μmol/l (10.0 mg/dl)
Q2. Case of mild renal insufficiency
Suspend ACEI and prescribe other antihypertensive drugs
Do not suspend ACEI and prescribe resin. Advise K restriction
Ethical considerations and statistical analysis
We sent a letter of intent that explained the importance of the survey and assured that we would publicize the results. We did not think that this study had ethical problems because it did not include any real patient data and all data were anonymous.
Differences in categorical data were analyzed with χ2 test. For continuous data, means were compared by unpaired t tests. Means and standard deviations are shown when appropriate.
Postal questionnaire to physicians and nephrologists
Characteristics of nephrologists and physicians
Nephrologist (n = 114)
Physician (n = 111)
18.4 ± 4.8*
15.6 ± 5.8
425,000 ± 544,000*
128,000 ± 273,000
Final training in nephrology (%)
As specialist, 88.6
As resident, 76.6
Experience of renal biopsy (%)
Experience of dialysis (%)
Daily renal patients
18 ± 19*
0.71 ± 1.8
Use serum creatinine for outpatient (%)
Use creatinine clearance for outpatient (%)
Audit of serum creatinine that would indicate referral to nephrologist
Audit of continuance of ACEI in the treatment of a case of renal insufficiency
As we all know, physicians are different from specialists in skill, knowledge and experience in a specialty. There may be no problem if differences between physicians and specialists are limited to a very specialized area, but serious problems would arise if there were differences between basic views held by physicians and specialists. However, there has been very little investigation of such differences between physicians and specialists.
The first author of this paper moved from the renal division to the department of general practice in the same medical school hospital 6 years ago and became aware of big differences between doctors working in these two areas. One notable difference was that physicians showed little interest in mild renal insufficiency and did not know how to use ACEIs appropriately for patients with mild renal failure. The author realized that this might be a serious problem if such differences existed nationwide.
Our results show that such differences are evident and critical in our country. Most physicians do not want to refer to nephrologists for mild renal insufficiency. Moreover, many physicians did not select similar treatment to nephrologists for mild renal insufficiency. This means that many Japanese physicians make different decision from nephrologists for renal insufficiency, which may worsen renal function.
Several criticisms: the adequacy of the sample, the adequacy of the evaluation of renal function, and the universality of the result, can be made of this study.
In Japan, selection of representative physicians for this study is difficult since there are few established training and certification systems for primary-care physicians in Japan. Graduates from one medical school in Japan were selected as representative general physicians for this study. The medical school was established 33 years ago with the aim of improving rural medicine, and all graduates of the medical school have an obligation to work in rural areas for a certain period. Most of them had training and clinical experience as general physicians. All of the physicians selected for this study were certified by the Japanese Society for Internal Medicine. The physicians selected for this study were therefore considered to be representative physicians in Japan.
There might be criticism concerning the use of serum creatinine as a marker of renal function in this study. In the guidelines for chronic kidney disease (CKD) published by the National Kidney Foundation, consultation and/or co-management with a kidney disease team is recommended for patients with stage 3 disease, and referral to a nephrologist is recommended for patients with stage 4 disease . Stage 3 is defined as creatinine clearance between 30 and 59 ml/min/1.73 m2 and stage 4 is defined as creatinine clearance between 15 and 29 ml/min/1.73 m2. The guideline for CKD treatment was also issued by the Japanese Society of Nephrology in 2007 . From the point of renal function, physicians are recommended to consult nephrologists for treatment of patients with GFR less than 50 ml/min/1.73 m2. However, as can be seen from Table 1, most Japanese doctors used serum creatinine as a marker of renal function in 2004. Only about one-third of physicians in Japan use creatinine clearance tests for outpatients. Moreover, many physicians work in areas with small populations compared with areas covered by nephrologists. The average age of inhabitants of small towns in Japan is generally much older than that of inhabitants of large cities. As shown by Cockcroft’s equation, renal function deteriorates with age, even if serum creatinine level does not change . Therefore, renal dysfunction might be considerably underestimated by physicians in Japan who use serum creatinine as a marker of renal function. In other words, the referral gap between nephrologists and physicians might be larger than shown based on figures for serum creatinine alone.
Although there are some limitations to our study, the results clearly show that there is a difference between decisions made by nephrologists and by general physicians in Japan regarding treatment of renal insufficiency. Results of previous studies also suggest that such a difference exists in other countries. Kalra et al. , a nephrologist, reported inappropriate use of ACEIs by general practitioners in the UK. De Lusignan et al.  reported that there have been many cases of overlooked renal insufficiency in the UK, based on results of analysis of data from medical databases. Akbari et al.  reported that the rate of detection of renal disease by family doctors was increased by 22.4% by using the equation for estimating creatinine clearance. These findings suggest a similar difference between nephrologists and physicians in other countries.
Recently, several studies have shown that serum cystatin C is a better indicator of renal function than not only serum creatinine but also creatinine clearance [14, 15]. However, the superiority of cystatin C over creatinine is controversial . Moreover, the use of serum cystatin C as a marker of renal function has not become widespread due to the high cost of the test and lack of information.
Classical powerful tools such as calculated creatinine clearance (CCr) and new renal markers such as cystatin C can alert physicians to the possibility of renal dysfunction. However, improvements in the sensitivity of methods for detecting renal dysfunction will be meaningless without a change in the perception of physicians. As we have shown here, perception of physicians is significantly different from that of nephrologists. Moreover, the recent CKD guidelines also state the importance of proteinuria and hematuria as referral condition to nephrologist: (1) spot urine protein-to-creatinine ratio 0.5 g/g or positive dipstick test of proteinuria (2+ or greater), (2) positive dipstick test (1+ or greater) of both proteinuria and hematuria. Thus, vigorous promotion of the CKD guidelines is essential. In conclusion, there is significant difference between decisions made by physicians and nephrologists regarding treatment for patients with serum creatinine concentration of 177 μmol/l. Further evaluation of the effectiveness of CKD guidelines is needed.
We are very grateful to Ms. Yoko Kasakura, Ms. Yuko Watanabe and Ms. Yukie Akutsu for secretarial support and to Ms. Waka Shibata for grammatical corrections. This study was supported by an intramural fund of the Division of Nephrology of Jichi Medical University.