C SRAT 10
C ..................................................................SRAT 20
C SRAT 30
C SUBROUTINE SRATE SRAT 40
C SRAT 50
C PURPOSE SRAT 60
C TO COMPUTE THE PROPORTION OF SUBJECTS SURVIVING, THE SRAT 70
C SURVIVAL RATES AND THE STANDARD ERRORS FOR SUCCESSIVELY SRAT 80
C REDUCED TIME PERIODS. THE SURVIVAL RATE IS COMPUTED FOR SRAT 90
C EACH OF K PERIODS, WHERE K IS A CONSTANT TO BE SPECIFIED ANDSRAT 100
C IS LESS THAN OR EQUAL TO N (WHERE N = TOTAL NUMBER OF SRAT 110
C PERIODS). SRAT 120
C SRAT 130
C USAGE SRAT 140
C CALL SRATE (N,K,X,IE) SRAT 150
C SRAT 160
C DESCRIPTION OF PARAMETERS SRAT 170
C N - THE TOTAL NUMBER OF PERIODS AFTER TREATMENT OR DIAGNOSISSRAT 180
C K - THE SPECIFIED PERIOD UP TO WHICH SURVIVAL RATES ARE TO SRAT 190
C BE CALCULATED SRAT 200
C X - AN INPUT AND OUTPUT MATRIX (N X 9) CONTAINING THE SRAT 210
C FOLLOWING INFORMATION SRAT 220
C FOR INPUT--STORED IN THE N ROWS OF EACH COLUMN SRAT 230
C COL 1 - NUMBER OF SUBJECTS ALIVE AT THE BEGINNING OF SRAT 240
C PERIOD SRAT 250
C COL 2 - NUMBER OF SUBJECTS WHICH DIED DURING THE SRAT 260
C PERIOD SRAT 270
C COL 3 - NUMBER OF SUBJECTS LOST TO FOLLOW-UP DURING SRAT 280
C THE PERIOD SRAT 290
C COL 4 - NUMBER OF SUBJECTS WITHDRAWN ALIVE DURING THESRAT 300
C PERIOD SRAT 310
C FOR OUTPUT--STORED IN THE FIRST K ROWS OF EACH SRAT 320
C COLUMN SRAT 330
C COL 5 - EFFECTIVE NUMBER EXPOSED TO THE RISK OF DYINGSRAT 340
C COL 6 - PROPORTION WHO DIED DURING THE PERIOD SRAT 350
C COL 7 - PROPORTION WHO SURVIVED DURING THE PERIOD SRAT 360
C COL 8 - SURVIAL RATE SRAT 370
C COL 9 - STANDARD ERROR OF THE SURVIVAL RATE SRAT 380
C IE- 1, IF K IS NOT IN THE CLOSED INTERVAL (0,N). SRAT 390
C 2, IF THE NUMBER OF SUBJECTS ALIVE AT THE BEGINNING SRAT 400
C OF PERIOD I IS LESS THAN THE SUM OF THOSE WHICH DIED, SRAT 410
C WERE LOST, OR WERE WITHDRAWN DURING PERIOD I (I=1,...N) SRAT 420
C 3, IF THE NUMBER OF SUBJECTS WHICH DIED, WERE LOST, OR SRAT 430
C WERE WITHDRAWN IN PERIOD I IS NOT EQUAL TO THE NUMBER SRAT 440
C ALIVE AT THE BEGINNING OF PERIOD I LESS THE NUMBER SRAT 450
C ALIVE AT THE BEGINNING OF PERIOD I + 1 (I=1,...N-1) SRAT 460
C SRAT 470
C REMARKS SRAT 480
C IF THE SUBJECTS IN A GIVEN GROUP ARE ALL DIAGNOSED OR SRAT 490
C TREATED AT THE SAME TIME, THE CONSTANT K MAY BE SET EQUAL TOSRAT 500
C N . IF THE SUBJECTS IN A GIVEN GOUP ENTER THE STUDY AT SRAT 510
C VARYING TIMES, K CAN BE NO GREATER THAN N-1. SRAT 520
C SRAT 530
C SUBROUTINES AND FUNCTION SUBPROGRAMS REQUIRED SRAT 540
C NONE SRAT 550
C SRAT 560
C METHOD SRAT 570
C REFER TO S. J. CUTLER AND F. EDERER 'MAXIMUM UTILIZATION OF SRAT 580
C THE LIFE TABLE METHOD IN ANALYZING SURVIVAL', JOURNAL OF SRAT 590
C CHRONIC DISEASES, DECEMBER, 1958. PP 699-712. SRAT 600
C SRAT 610
C ..................................................................SRAT 620
C SRAT 630
SUBROUTINE SRATE (N,K,X,IE) SRAT 640
C SRAT 650
DIMENSION X(1) SRAT 660
C SRAT 670
C INITIALIZATION AND ERROR CHECKING SRAT 680
C SRAT 690
IE=0 SRAT 700
NP4=4*N+1 SRAT 710
NP9=NP4+NP4+N-2 SRAT 720
DO 1 I=NP4,NP9 SRAT 730
1 X(I)=0.0 SRAT 740
IF (K) 2,2,3 SRAT 750
2 IE=1 SRAT 760
GO TO 45 SRAT 770
3 IF(K-N) 4,4,2 SRAT 780
4 DO 9 I=1,N SRAT 790
NP4=I+N SRAT 800
NP9=NP4+N SRAT 810
NP1=NP9+N SRAT 820
IF(INT(X(I)-X(NP4)-X(NP9)-X(NP1)+.01)) 5,6,6 SRAT 830
5 IE=2 SRAT 840
GO TO 45 SRAT 850
6 IF(I-N) 7,9,9 SRAT 860
7 IF (INT(X(I+1)-X(I)+X(NP4)+X(NP9)+X(NP1)+.01)) 8,9,8 SRAT 870
8 IE=3 SRAT 880
GO TO 45 SRAT 890
9 CONTINUE SRAT 900
15 L1=0 SRAT 910
L2=L1+N SRAT 920
L3=L2+N SRAT 930
L4=L3+N SRAT 940
L5=L4+N SRAT 950
L6=L5+N SRAT 960
L7=L6+N SRAT 970
L8=L7+N SRAT 980
L9=L8+N SRAT 990
LD=L2 SRAT1000
LE=L5 SRAT1010
LQ=L6 SRAT1020
SUM=0.0 SRAT1030
C SRAT1040
DO 40 I=1,K SRAT1050
C SRAT1060
C COMPUTE EFFECTIVE NUMBER EXPOSED TO RISK OF DYING SRAT1070
C SRAT1080
L1=L1+1 SRAT1090
L3=L3+1 SRAT1100
L4=L4+1 SRAT1110
L5=L5+1 SRAT1120
X(L5)=X(L1)-(X(L3)+X(L4))/2.0 SRAT1130
C SRAT1140
C COMPUTE PROPORTION OF DYING SRAT1150
C SRAT1160
L2=L2+1 SRAT1170
L6=L6+1 SRAT1180
X(L6)=X(L2)/X(L5) SRAT1190
C SRAT1200
C COMPUTE PROPORTION OF SURVIVING SRAT1210
C SRAT1220
L7=L7+1 SRAT1230
X(L7)=1.0-X(L6) SRAT1240
C SRAT1250
C COMPUTE SURVIVAL RATE SRAT1260
C SRAT1270
L8=L8+1 SRAT1280
IF (I-1) 20, 20, 25 SRAT1290
20 X(L8)=X(L7) SRAT1300
GO TO 30 SRAT1310
25 X(L8)=X(L8-1)*X(L7) SRAT1320
C SRAT1330
C COMPUTE STANDARD ERROR OF SURVIVAL RATE SRAT1340
C SRAT1350
30 L9=L9+1 SRAT1360
SUM=SUM+X(L6)/(X(L5)-X(L2)) SRAT1370
40 X(L9)=X(L8)*SQRT(SUM) SRAT1380
C SRAT1390
45 RETURN SRAT1400
END SRAT1410