140

The Journal of Health Science: April 2005, Vol. 7, No. 2. p140~p148

�� !"#$7 � 2 �

1�� !"#$%&'()*+,

2�� !"#$!%&'

3�� !"#$%&'()

�� !"2004 � 9 � 10��� !�"#2004 � 11 � 30��� !"#$2005 � 1 � 4 �

�� !"#$%&&'()*+,- 100 � 10 �

�� !"

�� !��� !�� !"#$%&'(

�� !"#$%&'(%)* +"#$%,

�� !"#$%&' 21�� !"#$%&'

�� !"#$%&�Diabetes Mellitus�� DM�

�� !"#$%&'()*+,-./012

�� !"#$% 10�� !"#$%&'()

�� !"#$%&'!()*+,-#. 1993

�� !"#$%&'()*+ ,- 2003 ��

�� !"#$%&'() 10,013�� !"�

�� 7.7%�� !"#$%&'(� !)*+

�� ! 44.4 ��� !"#$% 2003)�

�� !"#$%&'()&*+,-./0

�� !"#$ 2�� !"#$%&'(&)*

�� 6-12% (Chou, Li, & Tsai, 2001) �� !

�� !"#$%&'()*+,-+./01

�� !"

�� 1�� !

2�� !

3

��

�� �� !"#$%& (small area analysis)�� !"#$%&'()*+,-./01

�� !"#$%&'()*+,-(./0 1234�� �� !"#$%&'()*+,

�� !"#$%&'()*� !+,-./ 2000�� !"#$%&'()*+!,-./0

�� !"#$%&'()*+,-.%&/01234567+89+:;<=+>?@ABC

�� !"#$%&'()*�+,-./012)*�34*56/7%8$9:;<=>?*

�� !"#$%&'()*+*,-./0123456789�� �� !"�#$%&'

�� !"#$%&'()*+(,#$%-.$/0123456789:;<=4>?@*+

�� !"#$%&'(&)*+$%,-./�(&)*+$%0123456789:4;<

�� !"#$%&'()*+,-./01.2345678/9:;<=>?@ABCDEF

�� !"#$%&'()*+!,-./01234567!89�� �� !"# $%&

�� !"#$%&'()*+,-./0123456278$%'9:;<=>?@2ABC

�� !"#$%&'"()*+,-./0123456789!"$:;<=>?:;@AB

�� !"##$%&'()*+,-./01234(56789:;<=>?@AB6!"8

�� !"#$%&'()*+,-

�� !"#$%&'(%)*(%+,-./01

�� !"#$%&'()*+,-+./0123456

�� !"#$7 � 2 � 141

�� !"#$%&'()*+,-./ 0�

��� !"#"$%&'()*+,-���

� 4.3�� !"#$%&'((Chen, Yen, & Tung,

2001)�� !"#$%&'()*+,-./0

�� !��"#$%& '()*+,-./

�� !"#$%&'()*+,-./

� 1987�� !"#$%&'()*+,-

�� !"#$%&'(�)*+,-./01

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./012

�� !"#$%&$'()*+,-./�

�� !"#$��%&'! ()*+,-.

�� 1997�� !"#$%&#'()*+,-

�� !" (share care program)�� !"#$

�� !"#$%&'#$()*+,-./0

�� !"#$%&�� !"��� !"�

�� !"#$%&'() *+,-./01

�� !"#"$%&'()*+,-./01

�� !"#$%& 2000�� !"#$%&

��� !"#�� !"#

R. A. Spasoff (Robert, 1999)�� !"#$

�� !"#!$%&'()*+,-%./0

�� !"#$%&'()*+,-./0 !

�� !"#$%&%'"()*+,-.1996~2000

�� !"#�$%&'()*+,-./,0

�� !"#$% &'()*+,-./��

�� 2002��� !"#$%&'()*+��

�� �� !"#$�� !"�� !"#

�� !"#$%&'()*+,-./0-1

�� !"#$�� !"#$%&'�� !

�� !"#$%&'()*+,-./01(small

area analysis) (Akazawa, 1994; Philbin et al., 2003;

Richter & Berger, 2000)�� !"#$%&'(

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./!01

�� !"#$%&'()*+,-./012

�� !"

�� !(Chou, Chen, & Hsiao, 1992)��

�� !"#$%&' 30�� !"#$%&'

�� !"#$%&'(�)*�+, 12.4 % �

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+ !",-./'

�� !"#$%$&'(�� !"#$%&

�� !"#$%&�� !"#$%&'()*

�� !"#$%& '()*+,-./01

�� !"#$%&'($)*+,-./01

�� !"#$ (Chou, Chen, & Hsiao, 1992)�

�� !"#$%&'()*+,-./01

�� !"#$%&'()*+,-./012�

�� !"#$%&'(#)*+,-./01

��

�� !"#$

�� !"#$%&'()*+,-./01

�� !"#$%&'()�*+,-./01

��� !"�� 2000�� !"#$%&'(

�� !"#$%&'()*+,-./0��

�� !"#$%&'()*+,-./012

�� !"#$%�� 2000�� !"#$%&

�� !"#$%&'()*+,-.!"/0

�� !"# 1999�� !" 2001�� !"#$

��� !"#$%&''()*+ ICD-9-CM �

�� !"#$%&'�� ! 250�� !"�

�� ! 6480�� !"#$� 6488�� !"

�� !"#� 7751�� !"#$%� 7902��

�� !"#$%&� 7915�� !� 7916��

��� A-CODE � 4 �� A181�� !"��

�� !"#$%&'()*+,-./012

�� !�"#$ %�� 2003��� !"

�� !"#$%&'()*+,-."#$/

�� !"#

�� !"#$% &'()*+,-./0

�� !"#$%"&%'()*+,-&%'

�� !"#$%&'()*+,-�� !"

1986��� !"#$%&'(#)*+,-.

142

The Journal of Health Science: April 2005, Vol. 7, No. 2. p140~p148

�� !"#$7 � 2 �

�� !"#$%&'()&*� !"#+,

�� !"#$%&'()$%*+,&-./

�� !"#$%&'()* +,-./ +

�� !"#�$%&'()"*+,"-./

�� !"#$%&'()*+,-.!/01

�� !�"#$%�&'%�()%*%�+

�� !"#$%&'()! "�*+,-�

�� !"#$%&&'()*+,-�./0

�� !"#$%�� !�� !"#$%&

�� !"#$%&'()*+',-.'/0

�� !"#$%&'()*+,-./0123

�� !"#$%&'()*+,-./01

�� !"�#$%!&'()*+,-./&

�� !"�� !"#$%&'() !*+�

�� !"#$%&'

�� !"!#$%&'()*+,-./0

�� !"#$�� !"#$%� !&'$%

�� !"#�� !"#$%&'()*+,

�� !(Lorenz Curve)�� !"#$%&'(

�� !"#$%&"'()*+,-./01

�� !" (scatter plot)�� !" (correlation

coefficient)�� !"#$%&'()*+,-.

�� !"#$%&"#$'()*+,-./

�� !"#$%&'()*+, (hierarchical

poisson regression)�� !"#$%&'()*

�� !"#$%&'()*+,-./0.1

�� !"#$%

�� !"

�� ! 2000�� !!"#$%&'()

�� !"#$%&#'()*+,-*./ 2000

�� !"#$%&'()*�� !"#$%

�� !�� !"#$%&'()*+,-.

�� !"#$%&'()�� 2003���

�� 28.72�� !"#$%&' 27.20�� !

�� !"# 30.61�� !"#$%&'()*

�� !"#$%&'$(&)*+,-./0

�� !"#$%&'()*+,-.�� !

���� !"#$%&'()*+ 30�� !�

�� !"#�� !�� !"#$%&'

�� !"#$%&'()*#+,-./01

� 1�2000�� !"#$%&'()*+,-./012

�� !"#

�� !" �� � 1 � � 2 � � 3 � � 4 � � 5 � � 6 � � 7 � � 8 �

�� 28.72 29.47 26.43 26.19 26.81 29.24 31.21 34.04 37.73

�� .�� !"#$%&'()*+

�� !"#$(2003)�� !"#$%&'()*+,-.

�� !"

�� !"#$%&'()*+,-+./0123456

�� !"#$7 � 2 � 143

�� !"#$%#&'()*+,-.!/0

�� !"#$%&'()*+,-./0"#

�� !"#$%&'()*+,-./012

�� !"#$%&�� !�� !" 55 ��

�� !"#$ 54�� !"#$%&'()*

�� !"#$%&'()'*+, 45~54��

�� !"#$%&'()*+,�� !"�

�� �� !"#$%& '��() 55 ��

�� !"#$%&'()*+,- 65~74��

�� !"#$%&#'()*+,- 75��

�� !"#$%!&'()*+,-./&01

�� !"#$%&'()*+,- ./01

�� !"#$%&'()*+

�� !"#$%&'()*+,-./01

�� !"#$%&'()*+,-./012

�� !"#$%&'("#$)*'+,-

�� !"#$%&'()*+,-./012

(Lorenz Curve)�� !"#!$%&'()*+

�� !"#$%&'()*+,-./012

�� !�� !"#$%&'()*+�,-.

�� !"# $"%&'()*+,-./0

�� !"#$%&'()*+,-./'(0

�� !"#$%&'()*+,-./0123

�� !"#$%&'()*+,-./012

�� !"#$%&'()*$+,-./01

�� !"#$%&'()*+�� !",-.

35302520151050

malefemale

male

female27.42

1

27.2229.36

2

31.3427.73

3

30.7626.91

4

31.3326.79

5

31.4225.35

6

30.8724.84

7

30.7428.78

8

32.92

(0/0

0)

��.�2000�� !�"#$%&'()*+,

�� !"#$

��.�2000�� !"#$%&'()*+,�

��

0/00

1 2 3 4 5 6 7 8

180.00160.00140.00120.00100.0080.0060.0040.0020.000.00

0-1415-4445-5455-6465-7475

100

90

80

70

60

50

40

30

20

10

00 10 20 30 40 50 60 70 80 90 100

100

90

80

70

60

50

40

30

20

10

00 10 20 30 40 50 60 70 80 90 100

�� .�2000�� !"#$%&'()*+',-./0123

144

The Journal of Health Science: April 2005, Vol. 7, No. 2. p140~p148

�� !"#$7 � 2 �

�� !"#$%&"'()*+,-.�/0

�� !"#$%&'#()&*+,��!"

�� 2000�� !"#$%&'()*+',

�� !"#$%&

�� !"#$%&'()*+,-./01

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,�-./01

�� !"#$%&'()*+,-!"#$%

�� !"

�� !"#$#%&'()*+,-./0

�� !"#$%&'()*+,- !"#.

�� !"#$%&$'()*+,-./012

�� !"#$%&'()*+,-./012

�� !"#$

�� !"#$%&'()#$%*+(,-

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-. /01

�� !" (scatter plot)�� !" (correlation

coefficient)�� !"#$%&'($%&)*+

�� !"#$!%&'()*+!&,-�.

�� !"#$%&'( !)*+%&,-.

�� !"#$%�Spearman �� !"#$

0.74 � 0.58 � 0.43��� !"#$%&'()

Spearman �� !"#$��%& 0.78�� !

�� 0.23�� !"#$%&'()*+,-

�� !"#$%&'()*+,-./0,1

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./012

�� !"#$%&'()

�� !"#$%&'()*+,- (hierarchical

poisson regression)�� !"#$%&'()*

�� !"#$%&'()*+�� !��

�� 1~2�� !"#$%&'()*'+,-

�� !"#$%&'()*+, -./01

�� !"#$%&'()*�� 2�� !"

� Deviance � 29098.12�� 437.66��� !

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,!- 2�� 3

�� !"#$%&'()*+(,-.� !

�� !"#$%&'()*+,-./�difference

in deviance � (437.66 – 342.65) = 95.01 > chi-

square(7, .99) = 18.47��

�� 3�� !"#$%&'()* 4��

� 4�� ! 45~54�� !"#$%&'()*

�� !"#$%&'()*+,-. 45~54 ��

�� !"#$%&'()*+, 1.65�� !

4�� !"#$%&'()"*+,-./01

�� !"#$%& '(

�� ! 5�� 3�� !"#$%&'(

�� 5�� !"#$%&'()*+,-./

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./012

� 6�� !"#$%&'()#$%*+(,-

�� !"#$%&'(�Deviance/df = 1.7 �

�� 1�� !" overdispersion�� !��

�� 6�� !"#$%&'()*+,-./0

�� !"#$%&'()*+,-./01)

�� .�2000�� !"#$%&'()*+(

0 20 40 60 80

200

160

120

80

40

00 10 20 30 40 50

100

80

60

40

20

00 10 20 30 40 50

100

80

60

40

20

0

�� !"#$%&'()*+,-+./0123456

�� !"#$7 � 2 � 145

�� 2000�� !"#$%&'(

�� 1 �� 2 �� 3�� P � �� P � �� P �

�� -7.6383 <.0001 -12.8437 <.0001 -13.1021 <.0001��

���� 0.1083 <.0001 0.1946 <.0001 0.1928 <.00010-1415-44 2.0225 0.0005 2.0338 0.000545-54 4.3290 <.0001 4.3461 <.0001

�� 55-64 5.7306 <.0001 5.7453 <.000165-74 6.8063 <.0001 6.8225 <.000175+ 7.5849 <.0001 7.6085 <.0001�� �� 0.3226 <.0001�� 0.2475 <.0001

�� �� 0.2952 <.0001�� �� 0.2528 <.0001

�� 0.1946 <.0001�� 0.2970 <.0001�� 0.3441 <.0001

Criteria For Assessing Value Value/DF Value Value/DF Value Value/DFDeviance 29098.12 359.24 437.66 5.76 342.65 4.97Pearson Chi-Square 56544.88 698.08 439.91 5.79 346.44 5.02Log Likelihood 24893.02 39223.25 39270.76

�� 4 �� 5 �� 6 �� 7�� P � �� P � �� P � �� P �

�� -12.8208 <.0001 -13.7840 <.0001 -14.2577 <.0001 -13.8293 <.0001��

���� -0.2635 <.0001 0.2276 <.0001 -0.2456 <.0001 -0.2430 <.00010-1415-44 1.8418 0.0015 2.3423 <.0001 2.5591 <.0001 2.3401 <.000145-54 3.5592 <.0001 4.6602 <.0001 4.2364 <.0001 4.0401 <.0001

�� 55-64 3.9980 <.0001 5.9453 <.0001 4.3663 <.0001 4.2841 <.000165-74 4.4180 <.0001 6.8181 <.0001 4.3287 <.0001 4.4052 <.000175+ 5.6099 <.0001 7.2526 <.0001 4.8905 <.0001 5.1660 <.0001�� �� 0.2266 <.0001 0.1977 0.0001 -0.0305 0.5733�� 0.2401 <.0001 0.0846 0.1696 -0.0802 0.1992

�� �� 0.3392 <.0001 0.0500 0.5495 -0.1329 0.1142�� �� 0.3741 <.0001 0.0001 0.9990 -0.1018 0.2172

�� 0.4136 <.0001 -0.0691 0.4273 -0.0700 0.4169�� 0.5616 <.0001 0.0361 0.6740 0.0844 0.3223�� 0.5018 <.0001 0.0389 0.7058 -0.0803 0.4389�� 14.6338 <.0001 16.1531 <.0001 15.4182 <.0001�� 3.4256 0.0005 6.5579 <.0001 4.4298 <.0001

Criteria For Assessing Value Value/DF Value Value/DF Value Value/DFDeviance 158.52 2.33 330.44 4.86 114.02 1.70 139.16 1.91Pearson Chi-Square 155.63 2.29 330.47 4.86 112.77 1.68 138.40 1.90Log Likelihood 39362.82 39276.86 39385.07 39372.50

146

The Journal of Health Science: April 2005, Vol. 7, No. 2. p140~p148

�� !"#$7 � 2 �

� 4�� !"#$%&'()*+,-./01

�� !"#$%&'()*�+,-./01

�� !"#$%&'()*+�,-./ 7 ��

�� !"#$%&'()*+,-*./-0

�� !"#$%&'()*��+,-)./

� 7�� !"#$%&

�� !"

�� !"#$%&'()*+,-./01

�� !"#$%&'�� !"#$%"&�

�� !"#$%&'()*+,�$-./

�� !"#$%"&'()*+,-./0

�� !"#$%&'()*+$%,-./0

�� !"#$%&'()*+,-./0!1

�� !"#$%&'()*�+,-./01

�� !"#"$%!&"$'()*+,-.

�� !"#$%&'()*+,-$.%/01

�� !"#$%&'((utilization studies)

(McKinlay, 1992)�� !"#$%&'()*�

�� !"#$%&'#()*+,-./01

�� !�"#$%&'()*+,-.*/0

�� !"#$

�� !"�� !�� !"#$%��

�� �� !"#$%&'()*+,-./

�� !"#$%&'() *+,-#./0

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./0+1

�� !"#$%&'()*+,-./012

�� (small area analysis)�� !"#$%&'

�� !"#$%%&'()*+,-.

�� !"#�� !"#$�� !"#$

�� !"�� !"#$�� !"#$%&

�� !"#$%&'(%)(*+,-./0

�� !"#$%&'()*+,-./012

�� !"#$ %&'()*+,-./01

�� !"#$%&'()#*+&',-./

�� !"#$%&'()*+,-./(01

�� !"#$%&'()*%&+,-)%./

�� ! 2000�� !"#$�%&'()*+

�� !"#$%&'()*+,$-./01

�� !"#$%&'()*+,-./$0

�� 5�� 6�� !"#$%&'()*+

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-./012

��� !"#$%&'()*+,-./01

�� !"#$%&'()*+,-./01(

��� 2003��

�� !"#$%&'()*+,-./01

�� !"#�$%&'()*+,-./01

�� !"#$%&'()*+%&'()*,

�� !"#$%&'()*)+,-./01

���� !"#$%&'��� !"#$%

���� ! "#$%&'()*+,-./

�� !"#$%&'(�� 2003��� !

�� !"#$%&'()**+,-%&./

�� !"#$ (high intensity)�

�� !"#$%&'()*+,-$.�

�� !"#$%&'()*+,-./+,*

�� !"#$%&'()%*+,�-./0

�� !"#$%&'()*+,-./ 01

�� !"#$%&'()*+,-./012

�� !"#$%&'(��)*+,-./01

�� !"#$%&'"()*+,-./01

�� !"#$%&'()$*+,-./�0

��

�� !"

�� !"#$%&'()*+,-./01

�� �� !"#DOH91-NH-1011��� !

�� !"#$%&'()*+,#-./01

�� !"#$%&'()*�� !"#$%

�� !"#$%&'()*+,-./012

�� !"#$%&'()*+,-+./0123456

�� !"#$7 � 2 � 147

���� !"#$%&'()*

�� !"#

�� !"#(2003)�� !"#$%"#&'�

�� !"#$%&'

�� !"#$(1986)�� !"#$%&#'(

�� !"#$�� !"#$%� 12 � 287-

323 �

�� (2003)�� !"#$%&'()*+,-

�� !"#$%&'()*+,�� !"

�� !"#$%&'()*+,-./01

��

�� !"#$!%&'!%&'!()*!+

��(2002)� 1996~2000�� !"#$%&

�� !"�#�� !"#$%� 21 � 173-

180 �

Akazawa, Y. (1994). WHO Collaborating Centre

for Diabetes Treatment and Education. Diabe-

tes Res. Clin. Pract, 24(Suppl), S331-S333.

Chen, H. H., Yen, M. F., & Tung, T. H. (2001).

Computer simulation model for cost-effective-

ness analysis of mass screening for type 2

diabetes mellitus. A Diabetes Res Clin Pract,

39(suppl.1), S37-S42.

Chou, P., Chen, H. H., & Hsiao, K. J. (1992).

Community-based epidemiological study on

diabetes in Pu-Li, Taiwan. Diabetes Care, 15,

81-89.

Chou, P., Li, C. L., & Tsai, S. T. ( 2001). Epidemi-

ology of type 2 diabetes in Taiwan. Diabetes

Res Clin Pract, 39(suppl.1), S29-S35.

McKinlay, J. B. (1992). Health promotion through

healthy public policy: the contribution of

complementary research methods. [Review]

[53 refs]. Canadian Journal of Public Health.

Revue Canadienne de Sante Publique., 83

(Suppl.1), S11-S19.

Philbin, E. F. et al. (2000). The results of a random-

ized trial of a quality improvement interven-

tion in the care of patients with heart failure.

The MISCHF Study Investigators. Am. J. Med.,

109, 443-449.

Richter, B., & Berger, M. (2000). Randomized con-

trolled trials remain fundamental to clinical

decision making in Type II diabetes mellitus: a

comment to the debate on randomized con-

trol led tr ials (For debate) [corrected].

Diabetologia, 43, 254-258.

Robert, A. (1999). Spasoff�Epidemiologic Meth-

ods for Health Policy. Oxford University Press.

148

The Journal of Health Science: April 2005, Vol. 7, No. 2. p140~p148

�� !"#$7 � 2 �

Relationships between Morbidity and Mortality in Diabetes

and Socioeconomic Conditions in Taiwan

Shun-Chuan Chang1 Mei-Shu Lai

2 Ci-Yong Syu

3

Abstract

Objectives: This study applied small area analysis to examine regional differences for diabetes

morbidity, diagnosis, and mortality in Taiwan. It also investigated the association between such health

indicators and socioeconomic conditions in different districts in which patients live.

Methods: The database for diabetic patients was selected from two datasets: Outpatient claimed data

about detailed treatments from the National Health Insurance (NHI) as well as Inpatient claims data for

Year 2000. These datasets were cross-referenced with the residential addresses of diabetic patients. The

accuracy of all database for diabetic patients was assessed by previous study [11], and levels of

urbanization regions in which patients live were also analyzed. The Results in the development of small

area difference analysis for diabetes morbidity, diagnosis and mortality are derived from hierarchical

poisson regression.

Results: Lorenz curve results showed that the distributional heterogeneity of each level of urban-

ization might be dramatically stronger than those of diabetes morbidity level, i.e., the probable effect of

the socioeconomic conditions in different districts on diabetes mortality is more obvious than on diabetes

morbidity. The regression model further verified that while controlling sex, age, and prevalence factors

(morbidity), the estimation of the parameter at less urbanized area is larger than those of developed urban

areas, namely, the relative risk of diabetes mortality at less urbanized area is higher.

Conclusion: Morbidity and mortality are useful for evaluating the district’s basic medical demands.

In addition, diabetes inpatient data from different urbanized areas were adapted to investigate how

geographic-based utilization of diabetes inpatients affects its health status. Based on analytical results,

we concluded that while implementing diabetes health care measures, we must deliberate the different

needs between small areas in Taiwan to ensure that the medical resources allocation can be more rational

and feasible.

Key words: Diabetes mellitus, mortality, morbidity, small area analysis

1PHD student, The Institute of Business and Management, National Chiao Tung University.

2Professor, The Institute of Preventive Medicine, National Taiwan University.

3RA, Office of Statistics, Department of Health.

Received: Sep. 10, 2004 Revised: Nov. 30, 2004 Accepted: Jan. 4, 2005

Address Correspondence to: Shun-Chuan Chang 10F, 100, Aikuo. Rd., Taipei, Taiwan (R.O.C)