Graphs Reveal the Progression of Kidney Disease: A single measurement is not enough.
This article, and the rapid responses generated, discuss the significance of a “low” egfr calculated from a serum creatinine measurement in the elderly person, with the potential to diagnose a disease when none exists but also the potential, if a “low” egfr is ignored, to miss a condition that might respond to an intervention eg discontinuing a nephrotoxic medication, treating an infection or removing obstructing pathology.
All “abnormal” results may be indicating important pathology but also may be a “false” positive due to the reference ranges provided for elderly patients indicating “abnormality” in some without disease of clinical significance or with a kidney disease that is stable. When an egfr is below the reference ranges it is the clinician's responsibility to identify those with progressive disease of significance. The report of an “abnormal” value is, therefore, an indication to search for previous measurements of plasma creatinine and to generate a plan for sufficient further testing to enable an assessment of rate of progression. Repeat measurements may need to be immediate if insufficient previous values are available to aid assessment and when there are substantially changed values from the most recent measurements. Repeat measurement can often be planned to be collected at intervals of time in the future appropriate to the situation to generate an adequate sequence of data to assess if there is a decline or to establish stability.
Measurements should be considered as a sequence of measurements and not in isolation. The merits of considering sequential measurements of kidney function over time is not discussed in this review. They are critical to enable the clinician to assess if there is a rate of deterioration in excess of that anticipated with age. With sufficient sequential data an acceleration in a previous rate of deterioration, with a “change point”, can indicate a reason for deteroration eg introduction of a nephrotoxic medication.
Sequences of creatinine measurements are best presented as a graphic of the reciprocal of plasma creatinine or as the egfr (1). The importance of further investigations (eg measurement of urine albumin/creatinine ratio, examination of the urinary deposit and renal imaging) is increased when there is progressive change suggesting deterioration or an acceleration in deterioration not due to ageing. Graphical presentations should be provided by laboratory services or by clinic or hospital IT systems. Algorithms that analyse data statistically for trends in data and for change-points in sequential data, have been developed but are underused for assisting with the interpretation of sequential data and to provide alerts when the rate of deterioration changes. (2) (3)
(2) Knapp, MS, Smith, AFM, Trimble, IM, Pownall, R, Gordon, K (1983). Mathematical and statistical aids to evaluate data from renal patients. Kidney International, 24,474-486.
(3) Morgan EJ, Will EJ. Selection, presentation and interpretation of biochemical data in renal failure. Kidney International 1983, 24,438-445.
Rapid Response:
Graphs Reveal the Progression of Kidney Disease: A single measurement is not enough.
This article, and the rapid responses generated, discuss the significance of a “low” egfr calculated from a serum creatinine measurement in the elderly person, with the potential to diagnose a disease when none exists but also the potential, if a “low” egfr is ignored, to miss a condition that might respond to an intervention eg discontinuing a nephrotoxic medication, treating an infection or removing obstructing pathology.
All “abnormal” results may be indicating important pathology but also may be a “false” positive due to the reference ranges provided for elderly patients indicating “abnormality” in some without disease of clinical significance or with a kidney disease that is stable. When an egfr is below the reference ranges it is the clinician's responsibility to identify those with progressive disease of significance. The report of an “abnormal” value is, therefore, an indication to search for previous measurements of plasma creatinine and to generate a plan for sufficient further testing to enable an assessment of rate of progression. Repeat measurements may need to be immediate if insufficient previous values are available to aid assessment and when there are substantially changed values from the most recent measurements. Repeat measurement can often be planned to be collected at intervals of time in the future appropriate to the situation to generate an adequate sequence of data to assess if there is a decline or to establish stability.
Measurements should be considered as a sequence of measurements and not in isolation. The merits of considering sequential measurements of kidney function over time is not discussed in this review. They are critical to enable the clinician to assess if there is a rate of deterioration in excess of that anticipated with age. With sufficient sequential data an acceleration in a previous rate of deterioration, with a “change point”, can indicate a reason for deteroration eg introduction of a nephrotoxic medication.
Sequences of creatinine measurements are best presented as a graphic of the reciprocal of plasma creatinine or as the egfr (1). The importance of further investigations (eg measurement of urine albumin/creatinine ratio, examination of the urinary deposit and renal imaging) is increased when there is progressive change suggesting deterioration or an acceleration in deterioration not due to ageing. Graphical presentations should be provided by laboratory services or by clinic or hospital IT systems. Algorithms that analyse data statistically for trends in data and for change-points in sequential data, have been developed but are underused for assisting with the interpretation of sequential data and to provide alerts when the rate of deterioration changes. (2) (3)
(1) Rayner H, Thomas M, Milford D. (2016) "Plot all the Dots: Graphs Reveal the Progression of Kidney Disease", CHAP 3, Understanding Kidney Disease. Springer, London. http://www.theisn.org/education/education-topics/general-nephrology/item...
(2) Knapp, MS, Smith, AFM, Trimble, IM, Pownall, R, Gordon, K (1983). Mathematical and statistical aids to evaluate data from renal patients. Kidney International, 24,474-486.
(3) Morgan EJ, Will EJ. Selection, presentation and interpretation of biochemical data in renal failure. Kidney International 1983, 24,438-445.
Competing interests: No competing interests