Survival with an acute spinal-cord injury

Survival with an acute spinal-cord injury

002 I -96X I /79/0X) J Chron Dis. Vol. 32. pp. 269 283 0 Pergamon Press Ltd. 1979. Printed in Great SURVIVAL WITH SPINAL-CORD AN ACUTE INJURY ...

1MB Sizes 0 Downloads 38 Views

002 I -96X I /79/0X)

J Chron Dis. Vol. 32. pp. 269 283 0

Pergamon

Press Ltd.

1979. Printed

in Great

SURVIVAL WITH SPINAL-CORD

AN ACUTE INJURY

JESSF. KRAUS,* CHARLESE. NEMAT

Departments

0.

I -0269SO2.tx)/O

Britain

BORHANI

and

FRANTI,

RICHARD

S.

RIGGINS

of Community Health and Orthopaedic Surgery. School of Medicine. University of California. Davis. CA 95616. U.S.A.

Abstract--Survival of a cohort of 619 persons who had a spinal-cord injury while resident in one of I8 Northern California counties was determined according to external cause of the injury. level of impairment. age and sex. In addition. causes of death arc described and ohscrvcd and expected numbers of deaths are compared. The cast fatality rate was 51.7:‘{, over tivc six years after the spinal-cord injury. Probability of survival was lowest for those injured from motor-vehicle collisions and within this group. the chances of survival were poorest for pedcstrians and those involved in multiple-vehicle crashes. Quadriplegics had the lowest survival rates. Survival rates were higher for males than for fcmalcs. Mortality attributahlc to the spinalcord lesion was lowest in the young, increasing with age irrespective of type of external cause of the injury. severity of the spinal-cord lesion. or sex. The relative risk of death dccrcascd with age for each of thcsc parameters. Cardiorcspiratory complications wc’rc the most frequent immediate cause of late deaths for those surviving at least 45 days post-injury.

INTRODUCTlON

survival following injury is an important consideration in planning services for emergency treatment at the scene, transport of injured persons to treatment facilities, primary care to repair damages, and restoration of maximum functional capabilities. Such information is essential also in estimating the resultant economic impact. Since injuries to the spinal cord result in disabling damage. most often permanent, knowledge of immediate prognosis is important if strategies for treatment and restorative services are to be developed or expanded. Studies of survival have important implications also for the accuracy of reported mortality or case fatality rates associated with injury-related causes of death. That is. in many countries an injury-related death is not reported as such unless the death occurs within a specified period after the injury. This period varies considerably in different countries [ I23 making it very difficult, if not impossible, to compare rates between countries. This report describes survival in a cohort of persons with an acute spinal-cord injury. The cohort under study had sustained spinal-cord injury in 1970 or 1971 while resident in one of 18 Northern California counties. Survival is reported by the external cause of injury, level of impairment at time of hospitalization, age, and sex of the person injured. In addition, causes of deaths are described and observed and expected numbers of deaths are compared. LENGTH

of

METHODS

AND

MATERIALS

A spinal-cord injury was defined as an acute traumatic lesion of the spinal-cord. cauda equina. or nerve roots resulting in some degree of motor or sensory deficit or both. The injury must have occurred in 1970 or 1971 to a resident of one of the 18 *Requests for reprints should be sent to Dr. Jess F. Kraus. Department Medicine. University of California. Davis. CA 95616. U.S.A. 269

of Community

Health.

School

of

270

JESS F. KRAUS et ul.

prespecified Northern California counties,* regardless of place of occurrence or place of treatment of the injury. The population at risk is thus defined as the usual residents of the 18 Northern California counties. According to the United States Bureau of Census, the population of this area in 1970 was slightly more than 5.8 million persons (29.1% of the state’s population). Case ascertainment included an exhaustive systematic review of records, files, and reports from all hospitals and coroner’s offices in each of the 18 Northern California counties as well as from records of the State of California Departments of Health (Crippled Children’s Service), Rehabilitation, and Industrial Relations (Workmen’s Compensation). For completeness of case ascertainment, admissions records of major hospitals peripheral to the l&county region and major referral hospitals in Southern California, Western Nevada, and Southern Oregon were also reviewed. Autopsy and coroner records of the remaining 36 California counties were reviewed to identify those of the population at risk who died in California outside of the 18-county area. Data from all sources were abstracted by specially trained medical-record librarians who followed explicit abstracting protocols. Additional details of case ascertainment, data collection, and methods of quality control have been reported in detail elsewhere [7]. A list of persons (including age, date of birth, and social security number) with a spinal-cord lesion who were discharged alive from the hospital was submitted to the Vital Statistics Office of the State Department of Health for a statewide search of death records. Among the remaining living members of the cohort, individual patient follow-up after hospital discharge was attempted, for reasons of confidentiality, using indirect means. That is. current address and living status.were ascertained by searching telephone directories and contacting private physicians and utility companies. Finally. the current records of the California State Department of Motor Vehicles were examined to determine the patient’s current address or if the patient had moved out-of-state or had died during the previous 5-6 yr period. The period of patient follow-up was to December 31, 1976, 5-6 yr beyond the date of the spinal-cord injury in 1970 or 1971. All death certificates and autopsy records for identified cases were obtained and examined to determine whether the spinal-cord injury was related to the death directly or indirectly. If the spinal-cord lesion was noted in the death certificate and/or autopsy report to have contributed to the death, the death was included for certain analyses. Persons who died or were found dead at the scene of the injury were coded arbitrarily as having survived less than two minutes. For all others, dates and times of the injury and death, as recorded on the death certificate, autopsy protocol, hospital records, or official police report were accepted as reported. Six types of functional impairments diagnosed during hospitalization were utilized. These include: (1) quadriplegia (paralyses of all limbs-a complete lesion at C3 or above); (2) quadriparesis (paralyses in all limbs-complete or incomplete lesion C4-Tl or incomplete lesion Cl-Tl); (3) paraplegia (paralyses of lower extremities-complete lesion T2-L3); (4) paraparesis (partial paralyses-incomplete lesion at T2-L3); (5) other paralyses (including hemiplegia, Brown-Sequard syndrome, and central-cord syndrome); and (6) other deficit (including transient paresis, organ dysfunction). For certain analyses, the six types were grouped into complete lesions (quadriplegia-paraplegia) or incomplete lesions (quadriparesis-paraparesis and other deficits such as transient paresis, centralcord syndrome, etc.). For comparison purposes, Attributable Mortality Rates (AMR) were derived for specific subgroups of the cohort to measure the percentage of the deaths in excess of that normally expected in the members of the cohort during 5.5 years following *The counties included Shasta. Tehama. Nevada. Placer. and El Dorado (representing the mountain region of Northern California): Butte. Yuba. Sutter. Yolo. Sacramento. San Joaquin. and Solano (representing the central valley region of Northern California): and Contra Costa. Alameda. Marin. San Mateo. San Francisco. and Santa Clara (representing the coastal region of Northern California).

Survival

with an Acute Spinal-Cord

GEWERIL

I 70 - l I 0 I f605 : I- 50 z g % 40 -

ADJUSTED

____.,,\ALL,NJ”NE~pE . \ - 70

\

THOSE YIT” YASSW \Y"LTIPLE TRAUYA \ \

30 -

\

I

- 30

20 -

\ I I I

IO -

I

o_I

POPULATION-AGE

271

Injury

I,, I 2 3 5 MINUTES-

I, I, IO 15 30

- 20

_.

-10 -s__

I I I 2 4 HOURS-

--_ II 812

-_w_ I IICA I 2351 DAYS-

III 231 WEEKS*

II 23 MONTHS-

I 6

I ““l-0 I 2346 YEARS-

DURATION SURVIVAL t lo9 scale 1

FIG.

I. Percentage

survival

for persons with an acute spinal-cord

lesion.

the spinal-cord injury. Relative Mortality Ratios (RMR) are given also to illustrate the relative impact of spinal-cord injury on mortality compared to that expected in the general population as a whole. RESULTS

A total of 619 persons with a confirmed spinal-cord lesion which was incurred in 1970 or 1971 were identified among the usual residents of the 18-county region, producing an average annual incidence rate of 53.5 per million population. Of the 619 cases, 299 died at the time of injury or during hospitalization (235 died at the scene or were dead on arrival at the hospital and 64 died during hospitalization). Of the 320 persons known to be alive on discharge from the hospital living status was ascertained for all but one. Of those discharged alive, 35 had died by December 31, 1976, and 21 of these deaths* were related to the spinal-cord injury. Thus, the overall spinal-cord specific fatality rate was 51.7% [(299 + 21)/619]. Figure 1 illustrates results of life-table analysis of the data on follow-up of the cohort of 619 persons with an acute spinal-cord lesion. The period of follow-up ranged from five years (persons injured in 1971) to six years (injured in 1970). Fifteen percent of the original cohort of 619 died within two minutes of injury. Slightly more than 34’:1, died within the first hour. Within the first 24 hr. 40.5% died from their injuries, and 5.5 yr after the time of the injury 51.7% had died. In the general population the survival rate during this same period would be about 95%. Massive multiple trauma,t including damage to the spinal-cord was the cause of death for 94 of the 619 members of the chart. Figure 1 shows that 91% of those with massive multiple trauma died within the first hour, 96J)d,died within the first 24 hr. and none survived beyond one week. Since length of survival for these persons might unduly influence the overall pattern of survival for the entire cohort. life table analysis was repeated without this subgroup. As seen from the follow-up analysis (Table I). *Identified +Defined

c.I).32

3

by search of statewide

death records and individual

as extensive serious damage I

to multiple

patient

follow-up.

body parts. as in crushing

in.iuries

JESSF. KRAUS et al.

272

TABLE 1. SURVIVORSHIPFOR 525 PERSONSWITH AN ACUTESPINAL-CORULESIONEXCLUDINGTHOSEWITH MASSIVE MULTIPLETRAUMA-I8 NORTHERN CALIFORNIACOUNTIES1970-1971

Time after injury Minutes: Cl.9 24.9 5-9.9 10-19.9 2C29.9 30-59.9 Hours: I- 1.9 2-9.9 IO- 23.9 Days: I- 1.9 2-2.9 3-4.9 5-6.9 7-13.9 14-27.9 28.-59.9 W179.9 180-364.9 Years: l-l.9 2-2.9 3-3.9 44.9 5-5.9 6+

Persons alive at beginning of time interval

525 472 467 460 434 413 390 370 367 363 361 356 351 345 336 331 328 320 316 310 308 305 302 299

Number dying in interval*

Probability of dying in interval

Probability of surviving in interval

Percentage alive at beginning of interval

Standard error of the percentage

53t 5

0.101 0.010 0.015 0.056 0.048 0.056

0.899 0.990 0.985 0.944 0.952 0.944

100.0 89.9 89.0 87.7 82.8 78.8

1.3 1.4 1.4 1.7 1.8

0.05 1 0.008 0.011

0.949 0.992 0.989

74.4 70.6 70.0

1.9 2.0 2.0

0.006 0.014 0.014 0.017 0.026 0.015 0.009 0.024 0.012

0.994 0.986 0.986 0.983 0.974 0.985 0.99 1 0.976 0.988

69.2 68.8 67.9 66.9 65.8 64.1 63.1 62.5 61.0

2.0 2.0 2.0 2.1 2.1 2. I 2.1 2.1 2. I

0.019 0.006 0.010 0.010 0.010

0.98 I 0.994 0.990 0.990 0.990

60.2 59.0 58.7 58.1 57.5 57.0

2.1 2.1 2.2 2.2 2.2

26 21 23 20 4

5 5 6 9 5 3

8 A

6 2 3 3

*Excludes deaths unrelated to the spinal-cord injury. tlncludes persons found dead at the scene.

10.1% of those without multiple trauma died within the first two minutes. The probability of survival over the total period of follow-up was 0.570. Except where noted. persons with massive multiple trauma are excluded in the results which follow. As can be seen from Fig. 2, the probability of survival was lowest for those with a spinal-cord injury that resulted from motor-vehicle crashes or firearms. The probability of surviving a recreation-related* spinal-cord injury was 95% after 5-6 yr. Since 56% of the total cohort were injured in motor-vehicle collisions. survival of this particular subgroup deserves separate consideration. Figure 3 shows that 68% of injured pedestrians succumbed within the first hour and this group had the lowest overall probability (0.23) of surviving 5-6 yr after the injury. Except for the first hour following injury, the chance of survival was about 20% higher for a single-vehicle crash than for a multiple-vehicle crash. Persons involved in a motorcycle crash had the highest probability of surviving for 5.5 yr (0.70). Figure 4 illustrates percentage survival according to level of neurologic impairment irrespective of cause of injury. Those dead on arrival at the hospital or taken directly to the morgue were excluded, as well as those for whom no neurologic diagnosis was made (i.e., died in emergency room). Quadriplegics had the lowest survival rates 5-6 yr after the injury (30%). Over 70% of quadriparetics and those with other deficits such as transient paresis/organ dysfunction survived 5-6yr after the injury. About 84% of paraplegics, 96% of paraparetics, and 98% of those with other forms of paralysis (such as central-cord syndrome) survived 5-6yr after the injury. In general, 70% of those *Recreation-related spinal-cord injuries include those associated with athletics. sporting events. aquatic recreation. hiking (falls. etc.). Persons injured during recreational motor vehicle driving or hunting were classified as motor-vehicle collisions or firearms. respectively.

Survival

with an Acute Spinal-Cord

273

Injury

- 100

_-_____---_______----------~----_-w-m_

RECREATIOW

- 90

--4*Ls______._ l-__

- 60

--__

-.

P

‘-__

--__

--mm_

- 7og

F 70 -

h 2

& z

\

60 ; E ; 50 -

- --.‘.‘IL1:,.&.,,

-60, .. .. . . . - - - -::

3 Y - 50 L

.:*

4

1

30i

1 30

0

40

,.,,,,,,,,,,IIIIIIIII I 2 3 5 IO 15 30

I

2 4 HOURS-

MINUTES-

812

‘11’ 23L 23 WEEKS. MONTHS-

I

2351 DAYS-

lllll

L

6

0

2346 YEARS-

DURATION SURVIVAL I lo9 rcole) FIG. 2. Percentage

survival

persons with

an acute spinal-cord cause of injury.

for

lesion. according

IO external

with a complete lesion (quadri-paraplegics) were still living 5-6 yr post-injury, compared with 77% of those with an incomplete lesion (quadri-paraparetics). Survival patterns by sex are shown in Fig. 5. Except for the first 15-30 min. survival at each time interval was about 10% greater for males than for females. This finding was not consistent for the general types of lesions. As shown in Fig. 6. survival for 5-6 yr with a complete lesion was 54% for females and 75% for males. In contrast. with incomplete lesions, survival was poorer for males (76%) than for females (89%). Figure 7 illustrates the survival pattern by age. Sharp increases in deaths were evident from 10 min to one hour post-injury. more so for those less than 15 yr of age than for other ages. With that exception. the younger the person the greater the chances of surviving more than five years.

- 70

.......*. ......*_ 1

. . . .. ...... .. . .. . . . . ..... . . ... YULllPLl WaCLE -...

\

‘\

‘\

--A_‘_

503

--A___

3

“f ........

\ ---

‘\

i

___~LyrRIAM ---_

40 8

0

30

---.

---___

-_

-.-v-.-e-_ 20

ol,

I,,

I 235 MINUTES4

1,

1015

I

30

,

I

I

I 2 4 HOURS’

I,

I

812

I 2351 DAYS&

IIII

III

II

I

231 23 6 WEEKS+ MONTHS-

I

IIIII

I 2346 YEARS-

I,

DURATION SURVIVAL I lop scale 1 FIG. 3. Percentage

for persons with an acute spinal-cord

lesion by type of motor

vehicle

crash.

JFS F. KRAUSet al.

-...*

‘MR DEFICIT -._-.--..2::&_._

- 70 g - 60 ,’ 8 -50

‘; E E

-40n

- 30

4

j

20 0

1 III I 235 YINUTES-

II 1015

I 30

I1 II, I 2 4 812 HOURS -

I III, I 2351 DAYS-

III 1, 231 23 WEEKS-’ MONTHS-

I 6

20

I~~~~~~0 I 2346 YEARS-

DURATIONSURVIVAL

t lo9 SCOIC 1 FIG. 4.

Percentage

survival

for persons hospitalized with an acute spinal-cord to level of neurologic impairment.

lesion according

Figures 2. 4, 5, and 6 indicate that survival with a spinal-cord injury was related to age and sex of the patient as well as general type of the lesion and external cause of the injury. To assess the magnitude of these relationships, chi-square tests of association for certain defined subgroups of patients were applied. Attributable and relatiur ntorrality. The number of deaths expected annually in a group of 346 persons who had survived for at least a week following the injury was calculated from age-specific death rates given in the 1970 California Life Tables [ 111. The expected mortality was 2.71 per yr and, upon multiplying by 5.5 yr (assuming minimal effects due to aging in the cohort of 346 persons), the expected number of deaths was 14.90. The total number observed was 61, yielding a 5.5 yr mortality rate* attributable to the spinal-cord injury of 13.3% (Table 2). Excluding from the group of 61 all deaths

*Attributable

mortality

No. deaths observed -

rate =

No. expected deaths

x 100 in 5.5 yr.

No. persons at risk

- 100 -

90

- 80 ” f

YALES

- 70 $ It 2L 60 -

\ -mm

g - 60%

----.FEFLJS --__

-.

---__

50 -

E

----------_

\-

- 50

- 40

40 //

dl

1

‘I’

I 2 3 5 MINUTES-



IO I5



30



I





2 4 HOURS -



812

I’

“‘I

“I



2351 231 23 WEEKS+ MONTHSDAYS DURATION SURVIVAL



6



““’

Q

I 2 34 6 YEARS-

10

( IO9 scale 1 FIG. 5. Percentage

survival for persons with an acute spinal-cord

lesion according

to sex.

Survival

with an Acute Spinal-Cord

Injury

275 100

1 90

1

80

80

x - 70$ 5 -60: z Y -50%

1,

40

40

1 30

30 0

,,I

3 5 YINUTES-

I

2

,I

I1

I1

IO 15 30

2 4 HOURS-

I

II

I1111

812

23 OAYS-

III

I1

II

51 231 23 6 WEEKS’ MONTHS-

I

IIll

2346 YEARS-

I

0

DURATION SURVIVAL (109 scale) FIG. 6. Percentage survival

for persons with an acute spinal-cord

lesion by general impairment

level and sex. not related to spinal-cord trauma, the 5.5 yr AMR was 9.3%. As seen in Table 2, the age-specific AMR’s increased with age, irrespective of whether deaths not related to spinal-cord injury were considered or not. The number of deaths from all causes in each age group was significantly greater (P < 0.01) than expected. The number of deaths from spinal-cord related causes likewise was significantly greater (P < 0.01) than expected for all ages. Relative Mortality Ratios (RMR) given in Table 2 show that with advancing age, the ratio of observed to expected deaths decreases. Attributable Mortality Rates and Relative Mortality Ratios were higher in motor vehicle collisions than for all other external causes regardless of age for persons aged l-54 (Table 3). AMR’s were lowest in the young, increasing with age, whereas RMR decreased with age.. 100

1 90

-I

80

.. ......... . . . . . . . . . . .. . .

,,,,,

I

23 5 YINUTES-

1015 30

I 2 4 HOtJRS-

812

I 23 OAYS-

DURATION

51 231 23 6 WEEKS* MONTHS-

YEARS-

SURVIVAL

I lo9 scale ) FIG.

7.

Percentage

survival

for

persons

with an acute spinal-cord injury event.

lesion

by age at time

of

JRelative

from

deaths

deaths

No. persons

No. expected

(

No. observed

rate = >

RATES AND

RELATIVE

MORTALITY

RATIOS

FOR

significantly

greater

28.5 13.3

7.6 13.1

than

all causes

Attributable mortality rate (“;,)t

Deaths

expected,

2.4 4.1

17.6 7.6

P < 0.01.

Relative mortality ratio*

22* 47*

10* 15*

Number observed

Deaths

AT

LEAST

18.1 9.3

6.2 8.8

Attributable mortality rate (?$t

spinal-cord

PERSONS SURVIVING

18 NORTHERN CALIFORNIA TOUNTIES. 1970-1971

x 100 in 5.5 yr

deaths

28* 61*

12* 21*

Number observed

BY AGE.

MORTALITY

deaths

of observed

expected

at risk

(3). Number

Il.48 14.90

0.68 2.75

Number deaths expected 5.5 yr post-injury*

INJURY,

ATTRIRUTAHLE

A SPINAL-CORD

DEATHS.

observed-No.

Life Tables

CXPECTEI)

No. deaths ~

1970 Calif.

ratio =

mortality

mortality

tAttributab)e

deaths

58 346

55+ All ages

*Expected

149 139

Number alive 7 days post-injury

OF ORSERVEI~ AND

I 24 25 54

Age

TAHLE 2. NUMBER

FOLLOWING

1.9 3.2

14.7 5.4

Relative mortality ratio:

DAYS

causes

7

B 9 i*: P 5 z Q -_

mortality

ratio

mortality

deaths from

gAlI injuries from falls.

:Relative

tAttributablc

‘Expected

Motor vehicle All others

I 24 Motor vehicle All others 25-54 Motor vehicle All others 55+ Motor vehicle All other.+ All Ages

Age and cause of injury

=

(

of observed

>

x 100 in 5.5 yr.

deaths significantly

expected deaths

No. persons at risk

No. deaths observed --. --No. deaths expected

rate =

No. deaths observed--No.

(3). Number

greater

15.3 11.7

29. 32’

5.19 9.68

156 190

Life Tables

10.8 4.7 17.4 9.1 24.7 30.2

8. 4’ 13. 8, 8* 20.

0.32 0.35 1.32 I .42 3.56 7.91

71 78 67 72 18 40

1970 Calif.

Attributable mortality rate (%)t

Number observed

Number alive 7 days post-injury

24’ 23,

5.6 3.3

P -c 0.025.

8.0 4.7 14.4 3.6 19.1 17.7

6* 4+ II* 4 7’ 150

25.0 II.4 9.8 5.6 2.2 2.5

12.0 7.0

Attributable mortality rate (“/,)t

spinal-cord

Number observed

Deaths

causes

4.6 2.4

18.7 11.4 8.3 2.8 2.0 I .9

Relative mortality ratio:

A.r LUST 7 DAYS FOLLOWING

Relative mortality ratio:

than expected.

all causes

Number deaths expected 5.5 yr post-injury*

Deaths

TARLI, 3. NUMWR OF O~S~RVFD AND EXPI:CTI.I~DEATHS, ATTRIBUTABLE MORTALITY RATCS. AND RELATIVE MORTALITY RATIOS FOR PERSONS S[JRVIVING A SPINAL-CORI) INJURY. RY AGC ANU EXTERNAL CAUSE. I8 NORTHERN CALIFORNIA COUNTII~. 197W-1971

IlRelative mortality

ratio =

observed

expected

No. deaths

No. deaths

(3). Number

observed

deaths

3.36

103 243

quadri-paraplegics. quadri-paraparetics and all others. deaths from 1970 Calif. Life Tables

2.37 9.04

I2 46

*Includes tlncludes IExpected

13:

I .98

101

Incomplete 55+ Complete Incomplete All Ages Complete Incomplete

significantly

24: 37:

9: 191

8:

0.74

38

7: 5:

Number observed

Complete

II.51

KATIS

KLLATIVI:

all causes

greater

20.0 10.5

55.2 21.5

IO.9

19.1

12.8 4.8

than

Attributable mortality rate (‘z,)*

Deaths

P < 0.025.

7.1 3.2

3.8 2.1

6.6

10.8

29.2 11.4

21: 26:

9: 13

9::

6:

6: 4:

Number observed

1970-1971 Deaths

AT LEAST

17.1 6.0

55.2 8.5

7.0

13.8

IO.9 3.7

Attributable mortality rate (“,,)\\

spinal-cord

FOR PI:RSONS SL‘RVIVIN(;

UWNTIIS.

KATIOS

Relative mortality ratio

CALIFORNIA

MOHTALITY

expected.

18 NOKTHLKN

ANI)

TYIJI: OF LESION.

MORTALITY ANI)

0.24 0.44

ISY ~tii:

ATTKIIWTAI3LI.

INJURY.

Number deaths expected 5.5 yr post-injury:

A SPINAL-WRI)

1.XPI:cTI.I~ I)I:ATHS.

53 96

Number alive 7 days post-injury

ANI)

l-24 Complete* lncompletet 25-54

Age and type of lesion

TAISLI: 4. NUMIWK OF oIWKVI:I~

causes

7 l)Ays

6.3 2.3

3.8 1.4

4.5

8.1

25.0 9.1

Relative mortality ratio

FOLL()WING

%

zv) 2

LI E ?J f:

Survival with an Acute Spinal-Cord lnjury

279

As seen in Table 4, the number of deaths from all causes was significantly greater than expected for those with a complete lesion of the spinal cord (P < 0.025). The age-specific AMR’s and RMR’s indicated a greater risk of death with a complete lesion than with an incomplete lesion. A similar finding was made for deaths due to causes related to the spinal cord injury. As indicated earlier. percentage survival was higher for males than for females. It is not surprising then that when age-sex-specific numbers of expected deaths were considered, AMR’s and RMR’s were higher for females than for males for all three age groups for all causes of death and in two of the three age groups when only spinal-cordinjury-related deaths were included (Table 5). Im~diutr CUBISTS of‘ death. As mentioned earlier, 235 persons died at the scene of the injury or were dead-on-arrival at the hospital. The causes of death were immediate results of the injuries sustained (e.g.. hemorrhage, shock). Sixty-four deaths during hospitalization were related to the spinal-cord injury, and of the 35 additional deaths identified during 556 yr of follow-up, 21 were related to the spinal-cord lesion. Thirty-eight of those admitted to the hospital died within six days of the injury (a period arbitrarily selected as the period of emergency or primary treatment). Seventeen persons died 7-44 days after the injury, and 30 died 45-2163 days after injury. Numbers of deaths are summarized (Table 6) for general groups of immediate causes for three post-injury time periods (< 7, 7-44, 45 + days). For those dying less than seven days after the injury. the immediate causes (as listed on the death certificate/autopsy report) were about equally distributed among shock-hemorrhage, cardio-respiratory complications (including infection and/or embolism), traumatic disruption of the spinal-cord, and traumatic in.jury to the spinal-cord and other organs such as the brain. Almost all persons dying 7-44 days after the injury died from causes related to cardio-respiratory complications (including infection and/or embolism). The 30 deaths among persons surviving at least 45 days after the injury were related to cardio-respiratory or renal complications (including generalized sepsis and pyelonephritis). Seven deaths unrelated to spinal-cord trauma that occurred 45 days or more after the injury had cardiovascular causes, while six other deaths were from injuries sustained from motor-vehicle crashes (3 cases), suicide (1 case), or poisonings (2 cases).

DISCUSSION

Findings from this study are not strictly comparable with those of other published accounts [l-4. 6. 91 of survival in patients with spinal-cord injury. Some differences in study design between this present study and earlier studies are noteworthy. (1) Most of the earlier studies restricted the study groups to classically defined quadriparaparetics, whereas the present study included all patients with a diagnosed lesion of the spinal cord. (2) Mortality in previous reports was restricted to patients who had ‘recovered’ or had survived at least many weeks or even one year following spinal-cord injury, whereas in the present study duration of survival was determined from the time of the injury. (3) In other studies patient groups were limited to a series of cases seen in one hospital or a group of sitiilar hospitals (e.g., veterans hospitals) or to a subgroup of the population (e.g., veterans or those with casualty insurance). Patients in the current study constituted a cohort of injured persons arising from a defined population of 5.8 million people. (4) Some studies did not examine data on mortality in terms of the deaths expected to occur normally in the patient group during extended follow-up. In the present study. approximate Attributable Mortality Rates and Relative Mortality Ratios were calculated for that purpose. (5) The duration of follow-up in earlier reports varied from 10 to 24 yr. whereas mortality. in this present study, was determined for 5 to 6 yr of follow-up.

9* 3* 15* 6* 22s 6* 46* 15*

0.70 0.10

2.76 0.40

IO.17 2.50

13.62 3.01

II2 27

43 I5

269 II

Number observed

II4 35

Number alive 7 days post-injury

Number deaths expected 5.5 years post-injury*

12.0 15.6

27.5 23.3

IO.4 20.7

7.3 8.3

Attributable mortality rate (%)t

Deaths all causes

3.4 5.0

2.2 2.4

5.4 15.0

12.9 30.0

Relative mortality ratio:

No. deaths observed-No. expected deaths x 100 in 5.5 yr. No. persons at risk >

No. deaths expected

No. deaths observed

mortality rate =

SRelative mortality ratio =

tAttributable

20.5 3.3

l9f 3

7.9 II.7

8.3 5.6 20.7

3* 9* 6*

35* 12*

5.5

Attributable mortality rate (%)t

7*

Number observed

Deaths spinal-cord causes

2.6 4.0

1.9 I.2

30.0 3.3 15.0

10.0

Relative mortality ratio:

OF OBSERVEI)AND EXPECTEI)DEATHS. ATTRIIKJTARLI: MORTALITY RATES, ANI) RELATIVE MORTALITY RATIOS FOR PERSONS SURVIVING AT LEAST 7 DAYS FOLLOWING A SPINAL-CORD INJURY,BY AGE AND SR(, 18 NORTHERNCALIFORNIACOUNTIES,197Cb-1971

*Expected deaths from 1970 Calif. Life Tables (3). Number observed deaths significantly greater than expected, P < 0.025.

l-24 Males Females 25-54 Males Females 55+ Males Females All ages Males Females

Age and sex

TARLE 5. NUMBER

)z 7 z z E % %

Survival with an Acute Spinal-Cord Injury

281

TABLE 6. NUMBER OF DEATHS BY IMMEIJIATli CAUSE BY AGE AT TIME OF SPINAL-CORD INJURY FOR THREE TIME PERIODS FOLLOWING INJURY. 18 NORTHERN CALIFORNIA COUNTIES, 1970-1971

Age at time of spinal-cord injury Time period and immediate cause Deaths less than 7 days post-injury Shock- hemorrhage Cardio-respiratory* Trauma to spinal-cord only Trauma to spinal-cord and other organs All causes Deaths 7-44 days post-injury Cardio-respiratory’ Renal failure+ Trauma to spinal-cord All causes Deaths more than 44 days post-injury -spinal-cord related Cardio-respiratory* Renal failurct Trauma to organ system All causes Deaths more than 44 days post-injury -non-spinal-cord related Cardiovascular Neoplasm Other injuries1 All causes

25. 44

4 3 I 3

2

I

1 I 3

I I 2

2 4 6 3

9 9 9 II

II

7

5

I5

38

2 -. -. 2

4

5 I

4

6

3 _~ 2 5

I4 I 2 17

4 3 I 8

4 2

7 2 I IO

3 3

18 IO 2 30

2 _ 2 4

5 I

_ 3 3

45-64

All ages

~25

_

6

_ I I

65+

6

6

7 I 6 I4

‘Includes pulmonary infarction. embolism. and infection. tIncludes sepsis and pyeloncphritis. SInjuries sustained from external causes occurring after discharge from hospital following spinal-cord injury. i.e.. motor vehicle. suicide. aircraft. Despite the differences in study design described above, some findings of this present study are consistent with other published accounts. These include:

(1) poor (2) those (3)

Regardless of length of time of follow-up, persons with a spinal-cord injury show survival rates; Patients with complete lesions of the spinal cord show poorer survival rates than with incomplete lesions; and Risk of death is greatest in older age groups with a spinal-cord lesion.

One difference in results between this study and other published reports is noteworthy: the chief causes of death among those surviving at least 45 days post-injury in this study were related to cardio-respiratory complications (e.g., pneumonia, embolism) whereas in three earlier reports [ 1.3,6] the chief cause of death was renal failure. Almost half of the persons experiencing traumatic spinal-cord injury were alive 5-6 yr later. Of those who survived at least 24 hr. 84% were alive 5-6 yr later, a finding reasonably consistent with the figure of 77% reported by Burke and her colleagues [2] or the 88% figure reported by Kurtzke [S] in his excellent summary of survival data on spinal-cord injured patients. Survival following an acute spinal-cord injury depends on many factors, including patient age, extent and location of the spinal-cord lesion(s), extent and severity of other injuries, availability and adequacy of both emergency and primary treatment, and the nature of medical complications. As also reported by others [l-4, 6, 8, 93, we noted that duration of survival was related to age. More specifically, survival after 6 days was poorest for those 55yr of age and older. Mortality related to spinal-cord injury was significant in all age groups but more so for the oldest group. Besides age (which reflects differences in exposure to injury as well as physiologic responsiveness to trauma), explanations for differences in survival for persons injured from broad categories of external causes may relate to the variation in extent and/or

282

JESSF. KRAUS et al.

location of other injuries as well as the location and severity of the spinal-cord lesion. Quadriplegics had the poorest survival rate 5-6 yr after injury. Attributable Mortality Rates and Relative Mortality Ratios were highest for those with complete lesions of the spinal cord regardless of age. The high 5-6 yr survival rates (95%) associated with recreation-related spinal-cord injuries (752, aquatic diving injuries) possibly reflect traumatic energy exchanges which are smaller in magnitude and less concentrated in time and in their actions on body tissues. as well as the younger ages of the injured persons. Furthermore, the cohort of persons suffering spinal-cord trauma in recreation-related activities was small. Conversely. the relatively poor 5-6 yr survival from motor-vehicle collisions (47%) suggests that they may have sustained more severe injuries from a greater amount of energy delivered in more concentrated fashion. The range of forces generated in motor-vehicle impacts is considerable for various kinds of motor-vehicle collisions. For example, impact by an automobile to a pedestrian can deliver extensive damaging forces to a whole body area as well as the damage from secondary impacts with the ground. On the other hand, occupants in a motor-vehicle may sustain fewer and/or less extensive injuries because of the lesser energy exchange resulting from the ‘packaging’ of occupants within the vehicle. Motor-vehicle collisions (mostly single or multiple-vehicle crashes involving passenger cars) accounted for 56% of all spinal-cord injuries. The probability of death over more than five years (excluding those sustaining massive muitiple injuries) was 0.53 for persons involved in all types of motor-vehicle collisions, 0.40 for single-vehicle crashes, and 0.58 for multiple-vehicle crashes. The high risk of death from spinal-cord injury, particularly within the first 24 hr, indicates that major attention should be given to countermeasures to enhance chances for survival and recovery. Haddon [5] detailed 10 countermeasure strategies for reducing losses associated with energy exchanges damaging to the human body. Since a large majority of deaths related to spinal-cord injury involved motor-vehicle collisions, the strategies illustrated here focus on this aspect but are equally appropriate for other forms of mechanical energy or other energy exchange resulting in human death and injury. Discussion of the several concepts involving the pre-crash phase of motor-vehicle collisions are outside the scope of this study. In the motor-vehicle crash itself, one concept of prevention involves material barriers to separate the occupant from the energy being released. Crash padding is an example. A related method of prevention would be to eliminate or modify sharp edges, protruding obstacles, or other items that can inflict serious injury. Many automobile performance standards adopted over the last decade reflect this concept in changes in configuration and curvatures of instrument panels, etc. Another way is to modify the basic structure of the vehicle to better absorb and manage the transfer of energy that must be dissipated in a crash. Provision of emergency medical care at the scene of the crash and during timely transport to a treatment center is another important procedure for prevention, as well as a broad spectrum of treatment facilities needed to stabilize the injured victim and to provide for recovery and long-term rehabilitation. Given the high risk of death following spinal-cord damage, counter-measures to enhance the chances of survival are of major importance. These measures must include crash-worthiness and occupant protection in motor-vehicle design and the provision of services to increase the probabilities of survival following a motor-vehicle crash. As indicated earlier. reported mortality or case fatality rates may be difficult or impossible to compare because of the differences between countries in the rules used for including deaths in calculations of rates. For example, in the United States, a death from causes related to an injury within one year of that injury is included for purposes of calculation of death rates. whereas in Belgium, Italy, and Portugal only persons dead at the scene or dead on arrival at the hospital are counted [lo, 12, 131. In Spain, the period is 24 hr. in France three days, and in New Zealand and the United Kingdom the respective periods are 28 and 30 days [lo, 12, 133. Tonge ef al. Cl33 point out

Survival

with

an Acute

Spinal-Cord

injury

283

that although a period of 30 days would exclude many deaths that are ‘certainly due to the accident,’ evidence of actual survival times of the victims is needed to evaluate the precision of the rates. They concluded from their investigation that a 30-day cut-off period would exclude about 4% of the deaths. Among the 299 deaths ascribed to a spinal-cord injury from hospital or other health facility records at the time of data abstraction, two followed the date of injury by more than one year. These two deaths were included in the calculation of the overall death rate (25.8 per million population) reported previously [7]. Twenty-one additional deaths ascribed to the spinal-cord injury were identified through a search of death-record files of the State of California and individual patient follow-up. Fifteen of these 21 ‘late deaths’ occurred more than one year after the injury. The inclusion of these late deaths changes the overall mortality rate to 27.1 per million population. Nine of these 21 late deaths involved a motor-vehicle collision. Including these 9 increases the motorvehicle specific death rate from 17.5 to 18.4 per million population. The error in reported death rates by failure to include the late deaths is about 5?;, for both rates. Acknow/c,dgrr,lc,nf,s-We thank Dr. William Haddon. Jr.. President. and Dr. Leon S. Robertson. Senior Bchavioral Scientist. of the Insurance Institute for Highway Safety. Washington. DC.. for helpful suggestions in the preparation of this paper. This research was supported in part by a contract with the Insurance Institute for Highway

Safety.

Washington.

DC. REFERENCES

I. 2. 3. 4. 5. 6. 7. 8. 9. IO. I I. 12. 13.

Breithaupt DJ. Jousse AT. Wynn-Jones M: Late causes of death and life expectancy in paraplegia. CMAJ 85 : 73-77. 1961 Burke MH. Hicks AF. Robins M c’t ul.: Survival of patients with injuries of the spinal cord. JAMA 172: 121.-124. 1960 Freed MM. Bakst HJ, Barrie DL: Life expectancy. survival rates. and causes of death in civilian patients with spinal cord trauma. Arch Phys Med Rehab 47: 457 463. 1966 Guttman L: Spinal-Cord Injuries: Comprehensive Management and Research. Oxford: Blackwell Scientific Publications. pp. 625633. 1973 Haddon W. Jr.: Energy damage and the ten countermeasure strategies. J Trauma 13: 321 331. 1973 Jousse AT. Wynne-Jones M. Breithaupt DJ: A follow-up study of life expectancy and mortality in traumatic transverse myelitis. Canad Med Assoc J 98: 770 772. 1968 Kraus JF. Franti CE. Riggins RS et u/.: Incidence of traumatic spinal cord lesions. J Chron Dis 2X: 471492. 1975. Kurtzke JF: Epidemiology of spinal cord injury. Exp Neural 48 (No. 3, part 2): 163 236. 1975 Nyquist RH. Bors E: Mortality and survival in traumatic myelopathy during nineteen years. from 1946 to 1965. Parapiegia 5 : 22-48. 1967 Robertson JS. Tonge JI: Duration of survival in traffic accident fatalities. Med J Austr 2: 571 579. 1968 Schoen R. Collins M: Mortality by Cause. Life Tables for California. 1950 1970. California State Department of Health. 1973 Smeed RJ: The International Comparison of Accident Rates. Int Road Safety Traffic Rev I : 16. 1953 Tonge JI. Czechowic AS. Robertson JS: Survival after fatal traffic accidents. Lancet 2: 670. 1967