Use of clinical laboratory databases to enable early identification of patients at highest risk of developing endstage kidney disease Dr David Kennedy Dr Hugh Rayner Dr Jessie Raju Miss Kamaljit Chatha Chronic kidney disease (CKD) • CKD is common – est. 9% adults in England. • Prevalence increased in older people, those with diabetes and/or high blood pressure – upward trend. • Majority have mild to moderate disease – asymptomatic. • Minority progress to end-stage kidney disease (ESKD) and require kidney replacement therapy (KRT) (dialysis). • KRT = poor quality of life & costs £25K per patient per year. • Early intervention can delay or halt progression to ESKD. • Some patients remain undetected until very late. Birmingham Heartlands Hospital Good Hope Hospital Solihull Hospital eGFR test estimated Glomerular Filtration Rate (eGFR) • Calculated numerical result - marker of kidney function. • Based on serum creatinine conc. in blood. • Adjusted for age, gender and ethnicity. • From 2006 all UK biochemistry labs have reported eGFR for all creatinine tests requested for adults. • HEFT – approx. 9000 creatinine requests per week. • Results often looked at in isolation or compared to last 2-3. Objectives • Develop software capable of creating cumulative graphs of eGFR (up to 5 years data). • Create a system for identifying patients at highest risk of developing ESKD using data from the lab computer. • Build on previous HEFT diabetes renal system. • Monitor a large population (all clinics and community). • Clinical Scientists review eGFR graphs. • System must be capable of replication by other labs. HEFT Kidney Function Monitor • Oracle™ database updated daily with data from Heartlands and Good Hope lab computers • Generate lists of all patients from previous week • Aged 65 years or less with eGFR 50 or less • Aged > 65 years with eGFR 40 or less • Exclude renal patients and in-patients • Clinical Scientist reviews approx. 400 cumulative eGFR graphs identifying patients with significant declining trend or rapid deterioration. • High risk patients - report containing eGFR graph and information for further action sent to requesting doctor. Heart of England Kidney Function Monitor Patient Date of Birth Requesting Clinician Date/Time of Test ZZC00261281 17-Aug-1938 Dr K Arora 11-Apr-2012 04:13 Sex Location M Perry Park 291 Walsall Rd Estimated GFR 35 (ml/min) 01/01/2010 01/01/2011 100 90 80 70 60 50 40 30 20 10 0 01/01/2008 Est GFR (ml/min) 01/01/2009 GFR if Af ro Caribbean Other Tests From The Last 5 Years Test Highest 7.6 % HbA1c 5.1 mmol/l Serum K 144 mmol/l Serum Na 12.4 mmol/l Serum Urea 0.27 g/l Urine Protein Urine albumin:creatinine 30.5 mg/mmol ratio Marked as Inf orm Clinician Date 27/02/2012 11/04/2012 15/09/2008 11/04/2012 15/09/2010 15/09/2010 01/01/2012 Marked as Urgent Latest 7.6 % 5.1 mmol/l 134 mmol/l 12.4 mmol/l 0.07 g/l 1.7 mg/mmol Date 27/02/2012 11/04/2012 11/04/2012 11/04/2012 27/02/2012 27/02/2012 Diabetic Status on Renal Database Not Diabetic Dear Dr Arora, This results for this patient have reviewed as part of our departments system for monitoring chronic kidney disease using cumulative graphs of eGFR. Based on our criteria, agreed with the renal medicine department, this patient has been identified as at increased risk of progression to end-stage kidney disease. If you requirefurther advice or information for the management of this patient please contact Dr Hugh Rayner at Birmingham Heartlands Hospital. Email [email protected] for a copy of 'CKD Made Easy'. Please quote this patient's NHS number. Results – 1 Testing using historical data • Estimated 410 eGFR graphs to review per week. • Time to review graphs & generate reports approx. 3 hrs. • 15-20% of graphs reviewed by clinical scientists are flagged high risk. • Compared to the renal consultant - clinical scientists flag more patients as high risk but successfully identify those at highest risk. Results – 2 Testing using historical data • A random selection of patients were retrospectively flagged as high or low risk for one week in 2008. • Electronic data gathered in Jan / Feb 2012 (after 3.5 yrs). • All cause mortality was higher after 3.5 years in patients flagged as high risk compared to low risk. • The number of patients with a significantly declining eGFR over 3.5 years was higher for patients flagged as high risk compared to low risk. • The number of patients flagged at high risk who showed a significant decline in eGFR but had no evidence of specialist referral is estimated at up to 3% (780 per year). Estimated cost savings CKD progresses over years – showing early cost savings is thus impossible. A study at HEFT using cumulative eGFR graphs showed a significant fall in the number of diabetic patients requiring KRT after 5 year - estimated saving £390K Our monitoring system includes many more patients than the initial study therefore estimated savings are even more. Estimated cost of the new system at HEFT is £41K per year. If 20 patients over the next 5 years are detected earlier and KRT is delayed by a year net savings = £500K. 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 Number of diabetes patients starting KRT in HEKS Diabetes patients starting dialysis or transplanted per year 50 45 40 35 30 P<0.001 25 20 15 10 5 Rayner et al. BMJ Qual Saf 2011 0 Future plans • New system was introduced routinely at HEFT in April. • Quality data is being gathered prospectively. • Qualitative feedback (by questionnaire) of primary and secondary care clinicians will be collected. • Once embedded at HEFT, we plan to promote our new system through the clinical biochemist community and the West Midlands Renal Network. • We plan to extend the concept of cumulative monitoring of biochemical tests to other chronic diseases − Preparing Health Innovation Challenge bid. Conclusions • We have developed a system for lab staff to review cumulative eGFR graphs for a large population and identify patients at highest risk of developing ESKD. • We have tested the system using historical data and now introduced it into routine practice. • Reports with eGFR graphs are sent to clinicians highlighting patients at an earlier stage so that appropriate interventions to delay or halt deteriorating kidney function can happen earlier. • An smaller study at HEFT suggests this system may significantly reduce the number of patients needing KRT possibly saving £500K net after 5 years.