Forecast of the incidence, prevalence and burden of end-stage renal disease in Nanjing, China to the Year 2025

BMC Nephrology

Table 2 Goodness-of-fit statistical values

	Mean error^a	Maximum error	Minimum error	Mean percent error^b	Maximum percent error	Minimum percent error	R-square^c
Number of NJUEBMI population (No.)	2.27E-05	55591	−47633	−0.09	2.93	−2.88	0.997
Incidence (pmp)
All ESRD	3.23E-09	1.57	−1.55	−3.07E-03	0.73	−0.75	0.943
HD	−3.28E-09	0.94	−0.98	−1.73E-03	0.53	−0.54	0.971
PD	0.15	1.52	−0.65	0.56	5.66	−2.76	0.869
Kidney Transplant	−0.12	0.83	−1.04	−1.42	6.29	−10.68	0.756
Prevalence (pmp)
All ESRD	4.55E-09	58.32	−43.33	−0.07	5.36	−4.77	0.915
HD	1.17	24.73	−65.08	0.12	2.78	−8.16	0.931
PD	−2.73E-10	4.7	−4.92	0.41	10.93	−5.61	0.993
Kidney Transplant	−3.09E-09	10.31	−9.57	−0.06	6.42	−4.9	0.961
Health care costs (¥, in millions)
All ESRD	−0.64	16.61	−15.22	−1.57	7.09	−13.46	0.990
HD	−1.55E-09	8.89	−12.8	0.63	15.31	−12.41	0.991
PD	0.49	3.83	−1.61	5.62	37.36	−9.99	0.969
Kidney Transplant	−1.00E-10	0.95	−0.88	−0.36	29	−23.87	0.953
Per capita medical expenses (¥, in thousands)
All ESRD	−0.11	3.35	−2.89	−0.27	3.88	−4.48	0.942
HD	0.48	4.79	−4.6	0.45	5.34	−5.38	0.777
PD	−0.94	8.46	−8.69	−2.49	9.46	−13.08	0.852
Kidney Transplant	−2.50E-11	1.94	−1.34	−0.7	10.93	−12.1	0.917

Abbreviations: NJUEBMI Nanjing urban employee basic medical insurance, ESRD end stage renal disease, HD hemodialysis, PD peritoneal dialysis, pmp per million population
^aThe mean error indicates an average difference (2004 to 2014) of the forecasted values from the observed values. The maximum error indicates one of the 11 years (2004 to 2014) that exhibited the largest deviation of the forecasted value from the observed value, and the minimum error indicates the smallest deviation. A positive sign indicates over forecasting, whereas a negative sign indicates under forecasting
^bThe mean percent error reflects a proportional deviation of the mean error. Maximum and Minimum percent errors reflect proportional deviations of the largest and smallest errors, respectively
^cThe R-Square indicates the correlation between the observed values and the forecasted values

ISSN: 1471-2369