When laboratories report estimated glomerular filtration rates in addition to serum creatinines, nephrology consults increase
2009; Elsevier BV; Volume: 76; Issue: 3 Linguagem: Inglês
10.1038/ki.2009.158
ISSN1523-1755
AutoresArsh K. Jain, Ian K. McLeod, Cindy Yan Huo, Meaghan S. Cuerden, Ayub Akbari, Marcello Tonelli, Carl van Walraven, Rob R. Quinn, Brenda R. Hemmelgarn, Matthew J. Oliver, Ping Li, Amit X. Garg,
Tópico(s)Dialysis and Renal Disease Management
ResumoSerum creatinine alone can be difficult to interpret as a measure of kidney function such that chronic kidney disease might be under-recognized in the general population. In the province of Ontario, Canada, all outpatient laboratories now report estimated glomerular filtration rate (eGFR) in addition to serum creatinine. To determine the impact of this reporting on clinical practice, we linked health administrative data for more than 8 million adults of age 25 years or older over an almost 10-year period and conducted a population-based intervention analysis with seasonal time-series modeling to determine overall trends in the number and type of patients seen by nephrologists. Compared to the period when only serum creatinines were reported, the number of patients seen in consultation by nephrologists increased after eGFR reporting by an average of 24% (an absolute increase of 2.9 consults per 100,000 adults), an increase of about 23 consults per nephrologist per year. The greatest increases were seen in women (39% increase) and those 80 years of age and older (58% increase). Our study found that eGFR reporting was associated with a sudden increase in the number of nephrology consults. However, it remains to be seen whether the routine reporting of eGFR results in improved treatment and outcomes for those with chronic kidney disease. Serum creatinine alone can be difficult to interpret as a measure of kidney function such that chronic kidney disease might be under-recognized in the general population. In the province of Ontario, Canada, all outpatient laboratories now report estimated glomerular filtration rate (eGFR) in addition to serum creatinine. To determine the impact of this reporting on clinical practice, we linked health administrative data for more than 8 million adults of age 25 years or older over an almost 10-year period and conducted a population-based intervention analysis with seasonal time-series modeling to determine overall trends in the number and type of patients seen by nephrologists. Compared to the period when only serum creatinines were reported, the number of patients seen in consultation by nephrologists increased after eGFR reporting by an average of 24% (an absolute increase of 2.9 consults per 100,000 adults), an increase of about 23 consults per nephrologist per year. The greatest increases were seen in women (39% increase) and those 80 years of age and older (58% increase). Our study found that eGFR reporting was associated with a sudden increase in the number of nephrology consults. However, it remains to be seen whether the routine reporting of eGFR results in improved treatment and outcomes for those with chronic kidney disease. Chronic kidney disease (CKD) has been described as a global health concern.1.Tonelli M. Wiebe N. Culleton B. et al.Chronic kidney disease and mortality risk: a systematic review.J Am Soc Nephrol. 2006; 17: 2034-2047Crossref PubMed Scopus (1080) Google Scholar,2.Zhang Q.L. Rothenbacher D. Prevalence of chronic kidney disease in population-based studies: systematic review.BMC Public Health. 2008; 8: 117Crossref PubMed Scopus (672) Google Scholar A recent estimate suggests that about 8% of adults have CKD stage 3 or 4, when using the US National Kidney Foundation definition of CKD (an estimated glomerular filtration rate (eGFR) of 15–59 ml/min per 1.73 m2).3.Coresh J. Selvin E. Stevens L.A. et al.Prevalence of chronic kidney disease in the United States.JAMA. 2007; 298: 2038-2047Crossref PubMed Scopus (3662) Google Scholar Some health professionals do not recognize the presence of CKD. This may delay, or result in a complete loss, of an opportunity to counsel and treat patients, improve clinical outcomes, and reduce health-care costs.4.Jafar T.H. Stark P.C. Schmid C.H. et al.Progression of chronic kidney disease: the role of blood pressure control, proteinuria, and angiotensin-converting enzyme inhibition: a patient-level meta-analysis.Ann Intern Med. 2003; 139: 244-252Crossref PubMed Scopus (897) Google Scholar,5.Khan S. Amedia Jr, C.A. Economic burden of chronic kidney disease.J Eval Clin Pract. 2008; 14: 422-434Crossref PubMed Scopus (51) Google Scholar Although timely referral to a nephrologist or a specialized nephrology clinic is associated with improved patient outcomes, a significant proportion of patients requiring assessment by a nephrologist are not referred.6.Huisman R.M. The deadly risk of late referral.Nephrol Dial Transplant. 2004; 19: 2175-2180Crossref PubMed Scopus (71) Google Scholar, 7.Hemmelgarn B.R. Manns B.J. Zhang J. et al.Association between multidisciplinary care and survival for elderly patients with chronic kidney disease.J Am Soc Nephrol. 2007; 18: 993-999Crossref PubMed Scopus (132) Google Scholar, 8.Boulware L.E. Troll M.U. Jaar B.G. et al.Identification and referral of patients with progressive CKD: a national study.Am J Kidney Dis. 2006; 48: 192-204Abstract Full Text Full Text PDF PubMed Scopus (155) Google Scholar In clinical practice, the results of the serum creatinine test are used to estimate kidney function (GFR). However, serum creatinine may remain normal or only mildly elevated in the setting of CKD. Reliance upon serum creatinine alone has been suspected to lead to underdiagnosis of CKD.9.Hogg W. Rowan M.S. Lemelin J. et al.Why do family physicians fail to detect renal impairment?.Can Fam Physician. 2006; 52: 213Google Scholar,10.Hsu C.Y. Chertow G.M. Chronic renal confusion: insufficiency, failure, dysfunction, or disease.Am J Kidney Dis. 2000; 36: 415-418Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar The under-recognition of CKD is particularly prominent among women and the elderly.11.Swedko P.J. Clark H.D. Paramsothy K. et al.Serum creatinine is an inadequate screening test for renal failure in elderly patients.Arch Intern Med. 2003; 163 (Am Med Assoc. Ref Type: Generic): 356-360Crossref PubMed Scopus (257) Google Scholar,12.Duncan L. Heathcote J. Djurdjev O. et al.Screening for renal disease using serum creatinine: who are we missing?.Nephrol Dial Transplant. 2001; 16: 1042-1046Crossref PubMed Scopus (223) Google Scholar To improve CKD recognition, several international societies now recommend that kidney function be directly estimated with the use of an equation.13.Levey A.S. Eckardt K.U. Tsukamoto Y. et al.Definition and classification of chronic kidney disease: a position statement from Kidney Disease: Improving Global Outcomes (KDIGO).Kidney Int. 2005; 67: 2089-2100Abstract Full Text Full Text PDF PubMed Scopus (2202) Google Scholar, 14.Levin A. Mendelsohn D. Care and Referral of Adult Patients with Reduced Kidney Function. 2007http://www.csnscn.ca/local/files/CSN-Documents/CSN%20Postion%20Paper%20Sept2006.pdfGoogle Scholar, 15.National Kidney Foundation Kidney Disease Outcomes Quality Initiative. 2007http://www.kidney.org/professionals/KDOQI/guidelines_ckd/toc.htmGoogle Scholar Using the Modification of Diet in Renal Disease equation, a serum creatinine value along with the patient's age, sex, and race are used to calculate a better estimate of GFR than simply using the serum creatinine alone.16.Levey A.S. Bosch J.P. Lewis J.B. et al.A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation.Ann Intern Med. 1999; 130: 461-470Crossref PubMed Scopus (12277) Google Scholar The addition of such information to a laboratory report is termed 'eGFR reporting.' Estimated glomerular filtration rate reporting has been implemented in various regions around the world and has been the subject of intense debate.17.Centers for Disease Control and Prevention (CDC) Kidney disease mortality-Michigan, 1989–2005.MMWR Morb Mortal Wkly Rep. 2007; 56: 225-227PubMed Google Scholar,18.De Jong P.E. van der Velde M. Gansevoort R.T. et al.Screening for chronic kidney disease: where does Europe go?.Clin J Am Soc Nephrol. 2008; 3: 616-623Crossref PubMed Scopus (77) Google Scholar A recent survey confirms an increasing trend in the adoption of eGFR reporting among clinical laboratories; as of 2008, it was implemented in 70% of laboratories.19.Miller W.G. Reporting estimated GFR: a laboratory perspective.Am J Kidney Dis. 2008; 52: 645-648Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar Proponents believe that it improves the detection of early kidney disease.20.Levey A.S. Stevens L.A. Hostetter T. Automatic reporting of estimated glomerular filtration rate-just what the doctor ordered.Clin Chem. 2006; 52: 2188-2193Crossref PubMed Scopus (53) Google Scholar,21.Melamed M.L. Bauer C. Hostetter T.H. eGFR: is it ready for early identification of CKD?.Clin J Am Soc Nephrol. 2008; 3: 1569-1572Crossref PubMed Scopus (24) Google Scholar Others believe eGFR reporting may inappropriately label some patients who do not have a progressive pathological process as having CKD, simply because their eGFR is <60 ml/min per 1.73 m2. This could result in a large number of unnecessary consults to nephrologists, making eGFR reporting an overly sensitive 'screening test.'22.Rainey P.M. Automatic reporting of estimated glomerular filtration rate-jumping the gun?.Clin Chem. 2006; 52: 2184-2187Crossref PubMed Scopus (32) Google Scholar, 23.Clase C.M. Garg A.X. Kiberd B.A. Prevalence of low glomerular filtration rate in nondiabetic Americans: Third National Health and Nutrition Examination Survey (NHANES III).J Am Soc Nephrol. 2002; 13: 1338-1349Crossref PubMed Scopus (192) Google Scholar, 24.Glassock R.J. Winearls C. Screening for CKD with eGFR: doubts and dangers.Clin J Am Soc Nephrol. 2008; 3: 1563-1568Crossref PubMed Scopus (139) Google Scholar Regardless of one's position on eGFR reporting, its widespread use necessitates accurate assessments of its impact. Knowledge of the effects of outpatient eGFR reporting on the number of nephrology consults performed and the demographics of patients seen would inform this debate. It would help guide the adoption of this reporting strategy in other regions. Therefore, we conducted a population-based study to determine the effect of eGFR reporting on consults to nephrologists. The number of new outpatient consults seen by nephrologists in Ontario, which were non-urgent and where the patient had not previously been seen by a nephrologist, was about 1531 per month in the year before eGFR reporting. This meant that each nephrologist, on average, saw about 134 such consults in 2005. The number of 'new nephrology consults' from January 2003 through September 2007 are presented in Figure 1. A consistent seasonal variation was seen in the series, with nadirs coinciding with the traditional vacation periods during late summer and December. The number of patients seen in consultation increased after the introduction of eGFR reporting. The step function characterized this as an increase of 24% (95% confidence interval (CI) 16–31) or an absolute increase of 2.9 consults per 100,000 adult population per month (95% CI 2.5–3.4). This translated into a crude increase of about 3433 consults per year for Ontario (286 per month (95% CI 170–402)), or an average of about 23 consults per nephrologist per year. We considered the consult rate in different segments of the population. There was an increase in the consult rate after eGFR reporting for both women and men. The percent increase was greater in women compared with that in men (P<0.01). The percentage increase among women was 39% (95% CI 28–51) and the increase among men was 14% (95% CI 4–24). This translated to an increase in the absolute number of consults per month for women of 207 (95% CI 147–267) and 86 (95% CI 23–148) for men. There was a greater percent increase among the elderly (≥70 years old) compared with younger adults (<70 years old); P<0.001. We observed a graded increase in the consult rate after eGFR reporting among older patients aged 60–69, 70–79, and ≥80 years old, respectively (see Table 1). The largest percent increase was among patients aged 80 years old or more (an increase of 58%, 95% CI 35–80). Changes in the consultation rate after eGFR reporting were no different among non-diabetic patients compared with those in diabetic patients.Table 1Increase in consult rate for various age categoriesAge (years)Percentage increase in number of consults (95% CI)Change in total number of consults each month (95% CI)25–491.7 (−17.0 to 20.3)4.4 (−44.6 to 53.3)50–598.8 (−4.2 to 21.8)19.5 (−9.3 to 48.3)60–6925.6*P-value<0.0005. (12.1–39.0)58.0*P-value<0.0005. (27.4–88.6)70–7938.7*P-value<0.0005. (26.7–50.6)113.7*P-value<0.0005. (78.6–148.9)≥8057.8*P-value<0.0005. (35.3–80.2)78.0*P-value<0.0005. (47.7–108.3)CI, confidence interval.* P-value<0.0005. Open table in a new tab CI, confidence interval. The tracer outcome, dermatology consults, exhibited the same seasonal variations in rate as nephrology consults. eGFR reporting had no effect on dermatology consults (P=0.16). There were seasonal variations noted for the laboratory tests, which were assessed, with nadirs noted in the late summer and December time periods. Serum creatinine is presented as a graph representative of all laboratory tests (Figure 2). There was an increasing trend in the absolute number of serum creatinine tests performed over time. However, there was no significant effect of eGFR reporting on any of the laboratory tests (all P-values≥0.11). In this study, eGFR reporting was associated with a significant and sudden increase in the number of new nephrology consults. The greatest increase was noticed among women and the elderly. As expected, there was a short 3-month period after the initiation of eGFR reporting where there was no increase in the consult rate (representing the lag time it takes for the lab test to be ordered and completed, referral to be made, and consult to be seen). Some would conclude that outpatient eGFR reporting resulted in an increased awareness of CKD in the community, triggering an appropriate increase in referrals. This is supported by the results of some smaller studies25.Akbari A. Swedko P.J. Clark H.D. et al.Detection of chronic kidney disease with laboratory reporting of estimated glomerular filtration rate and an educational program.Arch Intern Med. 2004; 164: 1788-1792Crossref PubMed Scopus (124) Google Scholar, 26.Quartarolo J.M. Thoelke M. Schafers S.J. Reporting of estimated glomerular filtration rate: effect on physician recognition of chronic kidney disease and prescribing practices for elderly hospitalized patients.J Hosp Med. 2007; 2: 74-78Crossref PubMed Scopus (35) Google Scholar, 27.Wyatt C. Konduri V. Eng J. et al.Reporting of estimated GFR in the primary care clinic.Am J Kidney Dis. 2007; 49: 634-641Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar, 28.Aghaie-Jaladerany H. Cowell D. Geddes C.C. The early impact of the United Kingdom Chronic Kidney Disease (CKD) guidelines on the number of new attendances at renal clinics.Scott Med J. 2007; 52: 28-31Crossref PubMed Scopus (12) Google Scholar and by the fact that women and the elderly were differentially affected by eGFR reporting (the populations in which CKD is least likely to be recognized). The implementation of eGFR reporting occurred in the absence of a provincially mediated comprehensive education program, suggesting that reporting practices themselves led to the increase in consult rates. This may have triggered the increased awareness of CKD. However, others may question whether this increased rate of consultation truly reflects an increased recognition of CKD. Some believe that it is inappropriate to diagnose CKD in elderly individuals on the basis of an eGFR of 45–60 ml/min per 1.73 m2, because some of these people are unlikely to progressively lose kidney function.29.Froissart M. Rossert J. Jacquot C. et al.Predictive performance of the modification of diet in renal disease and Cockcroft–Gault equations for estimating renal function.J Am Soc Nephrol. 2005; 16: 763-773Crossref PubMed Scopus (677) Google Scholar,31.Grimes D.A. Schulz K.F. Uses and abuses of screening tests.Lancet. 2002; 359: 881-884Abstract Full Text Full Text PDF PubMed Scopus (370) Google Scholar Labeling these individuals as having CKD is incorrect because their decrease in GFR may simply represent a normal, age-related decline in kidney function. If eGFR reporting resulted in a large increase in inappropriate referrals, its utility as a screening test could be questioned.31.Grimes D.A. Schulz K.F. Uses and abuses of screening tests.Lancet. 2002; 359: 881-884Abstract Full Text Full Text PDF PubMed Scopus (370) Google Scholar Although examining the laboratory results of referred patients could help determine consult appropriateness as defined by various guidelines, only the fact that a test was performed is collected in administrative databases for Ontario. Thus, we could not determine the actual eGFR or serum creatinine level that initiated the consult, the rate of decline of kidney function, or other issues (for example, evidence of concurrent hypertension or proteinuria). Noble et al.32.Noble E. Johnson D.W. Gray N. et al.The impact of automated eGFR reporting and education on nephrology service referrals.Nephrol Dial Transplant. 2008; 23: 3845-3850Crossref PubMed Scopus (61) Google Scholar found that eGFR reporting was associated with both an increase in inappropriate referrals and an increase in appropriate referrals, as defined by the Kidney Check Australia Taskforce. As well, some could argue that any consult by a primary care physician to a specialist is appropriate, as it helps clarify a patient's diagnosis and future management. It is possible that the increased consult rate we observed may subside over time as primary care physicians become more familiar with eGFR reporting. This has been noted by others and may bring into question the increased awareness of CKD attributed to eGFR reporting.33.Richards N. Harris K. Whitfield M. et al.The impact of population-based identification of chronic kidney disease using estimated glomerular filtration rate (eGFR) reporting.Nephrol Dial Transplant. 2008; 23: 556-561Crossref PubMed Scopus (52) Google Scholar The limitations of this study need to be considered. We were not able to assess the number of referrals made to nephrologists or the reason for consultation. Instead, we could only determine that a consult was preformed. Consults seen by salaried physicians were not included; however, this represented only 5% of the physicians in Ontario. The appropriateness of referrals could not be assessed nor did we consider which physicians were referring the patients. Other contemporaneous events (for example, guidelines, educational efforts, and changes in billing practices) could have influenced the consult rate. However, when reviewing the literature and polling physicians across Ontario, we found no such publications or events that coincided with the dramatic and sudden increase in consultations observed after eGFR reporting. We could not assess the impact of eGFR reporting on wait times or if it resulted in a delay for nephrologists to see more urgent cases. Although we have attributed the changes to eGFR reporting, it may truly be the laboratory prompt that is causing these changes. When we considered all outpatient lab tests performed for the entire province of Ontario (about 8.9 million adults), the number did not change after eGFR reporting. This finding may be explained by three factors. First, the volume of tests performed in Ontario was much larger than the number of patients with CKD. For example, on a monthly basis, there were approximately 400,000 outpatient serum creatinine tests performed. If eGFR reporting increased testing in patients seen by nephrologists, or resulted in repeat testing by the primary care physician, this increase would still be a minor component of the absolute number of tests performed. Second, lab tests performed in hospitals were not captured in the provincial databases. Therefore, we were unable to assess the impact of eGFR reporting on hospital-based laboratory utilization. Finally, some tests, for example renal ultrasounds, do not have unique codes in Ontario health administrative data. Instead, similar tests were aggregated together at the billing level. In the case of renal ultrasounds, all types of abdominal ultrasounds were grouped together. Therefore, these types of tests were not assessed. We believe that the introduction of eGFR reporting was truly associated with an increase in the number of consults seen by nephrologists. To our knowledge, this is the first study to rigorously assess the impact of eGFR reporting on nephrologist consults in a large cohort of adults. Our data are population based and are comprehensive given the single-payer structure of health care in Ontario. Our exclusion criteria have effectively eliminated most non-routine consults, which would be seen by nephrologists (that is, urgent and emergency room consults). One of the concerns about eGFR reporting is that it could overwhelm nephrologists with a large number of consults.34.Coresh J. Eknoyan G. Levey A.S. Estimating the prevalence of low glomerular filtration rate requires attention to the creatinine assay calibration.J Am Soc Nephrol. 2002; 13: 2811-2812Crossref PubMed Scopus (118) Google Scholar The prevalence of CKD stage 3 or 4 is estimated at 8% or about 640,000 adults in Ontario. This raises the potential for large increases in the number of nephrology consultations if a substantial proportion of these were referred in response to eGFR reporting. However, there was only a modest increase of 3433 consults per year after eGFR reporting (23 per nephrologist per year), which seems unlikely to have a dramatic effect on wait times. Another concern about eGFR reporting is costs to the health-care system. We could not perform a comprehensive economic assessment; however, based solely on the number of 'first time' consults observed, the increase in nephrologist billings was about half a million dollars per year. An increase in laboratory tests could have great cost implications for the health-care system. In this study, eGFR reporting was not accompanied by an increase in the frequency of laboratory testing at the population level. However, it remains possible that the number of tests among those with CKD changed. The rationale behind eGFR reporting is that most patients with a low GFR will benefit from having been recognized and that ultimately this recognition will lead to improvement in patient outcomes, including the prevention of death and kidney failure. Studying such outcomes is difficult due to the prolonged time needed to identify them and their rare occurrence. During this time, other secular changes and medical advances would hamper an assessment of this intervention, particularly at the population level. The only true way to establish a causal relationship on such outcomes would be through a cluster randomized trial in which regions were randomized to receive eGFR reporting or not. However, such a study would offer a number of logistical challenges and the results would not be available for many years if ever. In the meantime, our assessment may help regions currently deliberating on whether to adopt eGFR reporting to anticipate changes in the utilization of health-care resources, before its adoption. Ontario is Canada's most populous province (38% of the Canadian population), currently with about 12 million residents, of which 8.9 million are adults. Ontarians have universal access to health care, including hospital care, physician visits, and laboratory services, through a single provincial government payer. In this structure, a consult is generated by a physician, usually a primary care physician, by sending a referral letter to a nephrologist. A referred patient is then booked and is subsequently seen by a nephrologist in consultation, and the nephrologist bills the government a pre-defined fee for their service (Ontario Health Insurance Program). During the study period (July 1999 to September 2007), about 67% of Ontarians were aged 25 years or older, with 12% exceeding 65 years of age; 51% were women; 6–9% of adults had diabetes;35.Lipscombe L.L. Hux J.E. Trends in diabetes prevalence, incidence, and mortality in Ontario, Canada 1995û2005: a population-based study.Lancet. 2007; 369: 750-756Abstract Full Text Full Text PDF PubMed Scopus (496) Google Scholar,36.Ontario Ministry of Finance Ontario Population Projections Update. 2007http://www.fin.gov.on.ca/english/economy/demographics/projections/2007/demog07.pdfGoogle Scholar 3.6% were African Canadians; and emigration was less than 1% per year (2001 Canadian Census). This study was conducted according to a pre-specified protocol, and ethics approval was obtained from the institutional review board at Sunnybrook Health Sciences Centre, Toronto, Canada. In January 2006, Ontario medical laboratories began adding eGFR results (using the abbreviated Modification of Diet in Renal Disease study equation37.Levey A.S. Greene T. Kusek J.W. et al.A simplified equation to predict glomerular filtration rate from serum creatinine.J Am Soc Nephrol. 2000; 11: 155AGoogle Scholar) to serum creatinines on their laboratory reports.38.OAML Communiqué Routine reporting of estimated glomerular filtration rate (eGFR). 2007http://oaml.com/PDF/FINAL%20DRAFT%20OAML%20eGFR%20Communique%20July%2012.pdfGoogle Scholar All serum creatinine reports had the patient's eGFR result and lab prompt appended to it. This change affected all adult outpatient tests. Hospital-based laboratories were not required to report eGFR. This change in reporting was implemented without a concurrent educational effort at the provincial level. All eGFRs were presented with a laboratory prompt that included an explanation of eGFR ranges (see Table 2). The same prompts were used regardless of age, sex, or race. Medical laboratories continued to report a serum creatinine level along with reference ranges.Table 2Typical eGFR reporting laboratory prompts for patients of varying degrees of kidney functionSource: personal communication with Gamma-Dynacare, one of the largest outpatient laboratories in Ontario.eGFR*eGFR measured by abbreviated MDRD equation in ml/min per 1.73m2.90–120Normal eGFR.eGFR*eGFR measured by abbreviated MDRD equation in ml/min per 1.73m2.60–89Slightly reduced eGFR is seen in approximately 30% of adults 20 years or older. Rule out kidney damage in those at high risk for chronic kidney disease.eGFR*eGFR measured by abbreviated MDRD equation in ml/min per 1.73m2.30–59Consistent with moderate chronic kidney disease if result confirmed by repeat measurement, with persistence for 3 months or more.eGFR*eGFR measured by abbreviated MDRD equation in ml/min per 1.73m2.15–29Consistent with severe chronic kidney disease.eGFR*eGFR measured by abbreviated MDRD equation in ml/min per 1.73m2.<15Consistent with kidney failure.eGFR, estimated glomerular filtration rate.All reports were followed by: for African Americans, the reported eGFR should be multiplied by a factor of 1.21 and re-interpreted accordingly.* eGFR measured by abbreviated MDRD equation in ml/min per 1.73 m2. Open table in a new tab eGFR, estimated glomerular filtration rate. All reports were followed by: for African Americans, the reported eGFR should be multiplied by a factor of 1.21 and re-interpreted accordingly. All outpatient laboratory tests and physician visits are recorded in the Ontario Health Insurance Program database. Each claim cites the patient, test or visit type, and service date. This database does not include claims submitted by salaried physicians (about 5% of health-care providers in Ontario). Importantly, there was no change in the percentage of physicians who were salaried during the study period. The Canadian Institute for Health Information Discharge Abstract Database (CIHI-DAD) records information about all Ontario hospitalizations. The Ontario Drug Benefit Database (ODBD) records all prescriptions filled to elderly Ontarians. The Registered Persons Database (RPD) records dates of birth, death, and place of residence. The Ontario Diabetes Database (ODD) identifies individuals in Ontario who have diabetes. Finally, the Institute for Clinical Evaluative Sciences (ICES) Physician Database (IPDB) records the specialty of all physicians practicing in Ontario. These databases were linked with encrypted, 10-digit health card numbers and have been extensively used in health-utilization and outcomes research.39.Alter D.A. Naylor C.D. Austin P. et al.Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction.N Engl J Med. 1999; 341: 1359Crossref PubMed Scopus (510) Google Scholar, 40.Juurlink D.N. Mamdani M. Kopp A. et al.Drug–drug interactions among elderly patients hospitalized for drug toxicity.JAMA. 2003; 289: 1652-1658Crossref PubMed Scopus (582) Google Scholar, 41.Juurlink D.N. Mamdani M.M. Lee D.S. et al.Rates of hyperkalemia after publication of the Randomized Aldactone Evaluation Study.N Engl J Med. 2004; 351: 543-551Crossref PubMed Scopus (1339) Google Scholar, 42.Mamdani M. Juurlink D.N. Lee D.S. et al.Cyclo-oxygenase-2 inhibitors versus non-selective non-steroidal anti-inflammatory drugs and congestive heart failure outcomes in elderly patients: a population-based cohort study.Lancet. 2004; 363: 1751-1756Abstract Full Text Full Text PDF PubMed Scopus (397) Google Scholar All Ontarians, at least 25 years old, were included in the study. We chose a cutoff age of 25 years to avoid consults from pediatric patients who are making the transition to adult nephrology care. We excluded individuals who did not have a valid health card number. On the basis of physician billings recorded in the Ontario Health Insurance Program database, people on chronic dialysis or those having undergone renal transplantation in the previous 4 years were also excluded. The primary study outcome was the number of new outpatient nephrologist consults seen each month that was non-urgent. We hypothesized that there would be a significant increase in the number of consults seen by nephrologists. A 'new nephrologist consult' was a consult claimed by a nephrologist, where the patient had not been billed for a nephrologist consult in the past 4 years (billing code A135). Non-urgent consults were defined as those consults that were not billed as urgent or from the emergency room. We did not assess the diagnostic codes for each consult as these data have not been validated. Nephrologists were defined as physicians who billed for chronic dialysis (for at least 10 weeks) and saw at least 10 new consults (A135) per year. Each year, a new list was generated to identify new and active nephrologists, as well as to remove retirees. This definition of nephrologist identified similar numbers of nephrologists in Ontario as described in the IPDB (a difference in number of less than 10% for years 1999–2005). Beginning in January 1999 and ending in December 2005, the total number of consults and their demographics were recorded on a monthly basis. There were 84 months of data before eGFR reporting was implemented, providing us adequate power for this study.43.McLeod A.I. Vingilis E. Power computations for intervention analysis.Technometrics. 2005; 47: 174-180Crossref PubMed Scopus (18) Google Scholar We then examined the monthly results from January 2006 until September 2007, giving us 21 months of post-implementation data. Temporally sequenced data are often autocorrelated and violate the assumption of independence, thereby negating the use of traditional regression.44.Box G.E.P. Tiao G.C. Intervention analysis with applications to economic and environmental problems.J Am Stat Assoc. 1975; 70: 70-79Crossref Scopus (1371) Google Scholar We therefore used time-series analysis to examine patterns in health-care utilization rates during the study period. The association between the change in the number of consults and eGFR reporting was assessed using interventional autoregressive integrated moving average (ARIMA) models with a step function and a lag of up to 3 months. This model allowed us to characterize the change in the number of consults from the pre- to post-eGFR reporting time periods. Autocorrelation, inverse autocorrelation, and partial autocorrelation functions confirmed model parameter appropriateness and seasonality. Autocorrelation functions and the augmented Dickey–Fuller test confirmed an appropriate trend over time (stationarity). Model adequacy was confirmed by examining the autocorrelations at various lags with use of the Ljung–Box χ2 statistic. Changes in crude, percentage, and standardized outcomes were compared with forecasted values.45.Hipel K.W. McLeod A.I. Time Series Modelling of Water Resources and Environmental Systems. 45. Elsevier Science, Amsterdam1994: 684Google Scholar Subgroups were pre-specified and analyzed by assessing the difference between parameter estimates of change (that is, intervention effect or step function) using a conservative pooled estimate of standard deviation, as the intervention effect followed a normal distribution.44.Box G.E.P. Tiao G.C. Intervention analysis with applications to economic and environmental problems.J Am Stat Assoc. 1975; 70: 70-79Crossref Scopus (1371) Google Scholar All P-values were two-sided and a value of less than 0.05 was considered significant. Statistics Canada provided census data on the number of people within each sex and age category in Ontario on 1 July of each study year. To determine the population for each time interval (month), linear extrapolation was applied between 1 July estimates for successive years. Crude and age- and sex-standardized rates for nephrology consults were calculated and analyzed. For the primary outcome, age- and sex-standardized rates were used to remove any changes that may have resulted from changes in the population structure over time.46.van Walraven C. Goel V. Chan B. Effect of population-based interventions on laboratory utilization: a time-series analysis.JAMA. 1998; 280: 2028-2033Crossref PubMed Scopus (128) Google Scholar For laboratory tests, crude totals on a monthly basis were tabulated. All analyses were performed using R for Windows, Version 2.6.1.47.R Development Core Team (2008) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria2008http://www.R-project.org/Google Scholar With regard to nephrology consults, we expected that eGFR reporting would differentially impact the number of consults among women and the elderly. Kidney function is routinely overestimated in these patients when using serum creatinine alone; with a marked reduction in GFR, there is often only a modest increase in serum creatinine. We believed that there would be a greater percent change in the number of consults for women and for the elderly (≥70 years old) compared with men and younger adults (<70 years old), respectively. As well, we hypothesized that there would be a differential effect on diabetics compared with non-diabetics. Diabetics have routine kidney-related screening (for example, serum creatinine and microalbumin to creatinine ratio). Therefore, eGFR reporting may not have as great an impact on them as it does on non-diabetics. We conducted several additional analyses for this study. First, as a control or tracer outcome for the primary outcome, dermatology consults were examined in a similar fashion. It was expected that dermatology consults would not be influenced by eGFR reporting, but could be affected by other secular health-care factors should they exist. Second, we examined trends in the use of outpatient diagnostic tests associated with kidney disease. We assessed the absolute number of outpatient tests performed in Ontario for serum creatinine, serum urea, renal biopsy, timed urine collections for creatinine clearance, urinalysis, and nuclear medicine GFR. We included all outpatient tests ordered for the adult population of Ontario. We hypothesized that eGFR reporting would increase the number of tests that were performed. All the authors declared no competing interests. We thank members of the Division of Nephrology, University of Western Ontario for their helpful comments. Grant support for the study was provided by the Canadian Institutes of Health Research. Dr Jain is enrolled in the Clinician Investigator Program at the University of Western Ontario. He receives funding from the Ministry of Health. Dr Quinn is supported by a Canadian Institutes of Health Research, Institute for Health Services and Policy Research Fellowship. Dr Garg is supported by a Clinician Scientist Award from the Canadian Institutes of Health Research. The Institute for Clinical Evaluative Sciences receives funding from the Ontario Ministry of Health and Long-term Care. The opinions, results, and conclusions reported in this paper are those of the authors, and are independent from the funding sources.
Referência(s)