Does Patient Cognition Predict Time Off From Work After Life-Threatening Accidents?

To our knowledge, this is the first study that used a representative group of severely injured accident victims to investigate the determinants of the number of days of leave taken because of the accident. Our aim was to collect as homogeneous a group as possible, with patients free from mental disturbances that were attributable to severe head injuries. The mean Injury Severity Score (22.1) and the Glasgow Coma Scale score (14.4) in our group indicated that this goal was achieved. Furthermore, the patients were excluded from the study if they showed any signs of previous psychological problems. The exclusion of patients who had attempted suicide or had been exposed to a physical assault further contributed to the homogeneity of the group.

Contrary to other groups that were drawn from accident victims seeking treatment for their posttraumatic psychological problems (22–24) or claiming compensation for personal injury (7), our group was collected consecutively, with the intensive care unit of the traumatology department of the University Hospital of Zurich as the single source. Only 10% of the eligible patients refused to participate in our study, an unusually low refusal rate compared with those in other published studies of accident victims (25–27). The dropout rate was also low. Comparative analyses of refusers and dropouts showed that our final group was representative of the population we intended to study. With regard to sociodemographic variables, our group was typical of groups of accident victims in general. Patients in this study were predominantly young men; marital status, living arrangements, educational level, and employment status corresponded to the sex and age distribution of admissions to our trauma surgeons’ intensive care unit. However, generalizability of our data to trauma populations with different socioeconomic and ethnic composition remains unclear.

This study has a number of limitations. First, the patients had to be excluded from the study if they did not speak German sufficiently. Proficiency in the official language of a country is a strong determinant of social integration; according to our clinical experience, patients with poor social integration have greater-than-average difficulties in dealing with the consequences of their accident. Therefore, in future studies, patients whose mother tongue is other than the country’s official language should be included by using interpreters. Second, the study did not include the full spectrum of accident victims, so inferences can only be made about those with severe accidental injuries. Also, certain types of injury are more likely to be related to unfitness for work, independent of the formal measure of the Injury Severity Score. For instance, patients with spinal cord injuries (Injury Severity Score=9) were not included in our study unless they had sustained additional injuries. Third, the patients’ self-assessments were not evaluated by using standardized instruments, but rather simple Likert scales. While the questions used have not been comprehensively validated, they certainly have face validity and their simplicity allows their use in routine clinical work. Furthermore, our results clearly support the predictive validity of the questions used. Finally, from our study design, one cannot be confident that the relationship between the attribution questions and outcome was a direct one, although this seems likely. Responses to these questions correlated with SCL-90-R scores (General Symptomatic Index by appraisal of accident severity [r=0.22, p<0.05] and by appraisal of coping abilities [r=–0.29, p<0.01]). As with other patient groups in the general hospital, a considerable proportion of our patients were likely to have been depressed, and depression in turn was likely to influence attributions (28, 29). However, despite each self-assessment correlating with the General Symptomatic Index, they were independent of each other. If they were both manifestations of depression, surely this would not be the case. Also, patients’ reports may have been influenced by the recency of the traumatic event, pain medication, and sleep medication. Nevertheless, whether the attributions studied are directly relevant to outcome or represent epiphenomena of depression or other posttraumatic psychopathology, they appear to be markers of outcome which, because of their simplicity, are potentially useful clinically and for research.

The key finding of this study was that patients’ appraisals of the severity of their accident and their ability to cope with the accident and its job-related consequences made substantial contributions to the duration of their time off work, independent of the severity of the injury. In addition, those who sustained their injuries during sports or leisure pursuits were likely to return to work sooner than those with injuries from other sorts of accidents, irrespective of the objective severity of the injuries.

The modest contribution of injury severity (assessed by the Injury Severity Score) to days of sick leave may have been due at least in part to the selection in this study of only patients with severe injuries. Had patients with mild or moderate injuries been included (thus covering the full range of Injury Severity Scores), the Injury Severity Score might have been a more powerful predictor of time taken off work. It is also noteworthy that, contrary to previous reports (6, 7), patients’ time taken off work was not independently predicted by their Impact of Event Scale scores. At the univariate level, Impact of Event Scale score correlated significantly with the number of days of leave taken; we assumed that in the multivariate analysis, the predictive power of the Impact of Event Scale score was suppressed by the patients’ appraisal variables, which each showed significant univariate correlations with Impact of Event Scale intrusion scores (Table 3).

In the present study, the strongest predictors of time taken off work were the patients’ self-assessments of accident severity and of their ability to cope. That each of these was apparently independent of the objective severity of injury is consistent with other work on patient appraisal of illness (28, 29). Cornes identified “psychological problems” as an important predictor of time taken off work because of accidental injuries (7) but without specifying these psychological problems in greater detail. In another study, objective medical ratings of loss of earning capacity were reported not to be predictive of overall occupational or psychosocial adjustment (30).

Patients who sustained their injuries during sports or leisure pursuits resumed work earlier than those whose accidents had other causes. To our knowledge, this finding has not been reported previously, although other authors have investigated samples of patients with different types of accidents (31–33). Several factors may contribute to this finding. First, patients in this group were more positive in their appraisals of their coping abilities than patients with other types of injury. Second, and possibly related to the first factor, people who have sustained injuries during activities that they freely choose to pursue are less likely to blame others for their accidents than those who were injured at work or in traffic accidents. It is possible that those injured during sports activities perceive such injuries as risks they choose to take, whereas work or traffic accidents are unlikely to be viewed in this light. Attributing blame to others has been associated in a variety of situations with unfavorable outcomes, although their influence on the outcome of accepting blame oneself has been less consistent (28). Third, individuals who have been physically active might be expected to recover more rapidly than other people from their injuries, although the duration of hospitalization did not differ significantly across the types of accident.

The influence of patients’ self-assessments on the number of days of leave taken that were attributable to the accident was not just statistically significant but makes a remarkable difference from a clinical and economic perspective, too. While Injury Severity Scores did not differ significantly among the four subgroups formed based on patient self-assessments, time taken off work varied between 4 and 9 months (Figure 1). Clearly, the results of this study could have numerous important clinical and therapeutic consequences, if they can be replicated in future work. First, attention to the cause of the trauma might assist in identifying subgroups of patients more likely than others to have longer time off work. Second, asking patients two brief and easily understandable questions may yield important information about key factors influencing the time to return to work. Third, our results suggest that an important focus for clinicians involved in the aftercare of patients with traumatic injuries should be to support and enhance their patients’ physical and psychological resources to cope with the accident and its sequelae. Finally, where a patient’s appraisals are pessimistic or negative, these are likely to be amenable to change. Under such circumstances, interventions might be helpful that aim to modify attributions, such as antidepressant treatment (if depression is present and is likely to have influenced the attributions) or cognitive behavior therapy. It has recently been reported that after myocardial infarction, therapeutically induced positive changes of illness perception were correlated with a faster return to work (34). The preventive effect of such interventions in accident victims, given at an early stage of recovery, and particularly the extent to which they could reduce the time to return to work, will need to be examined in randomized controlled trials.

Presented at the Third World Conference of the International Society for Traumatic Stress Studies, Melbourne, Australia, March 16–19, 2000; the 23rd European Conference on Psychosomatic Research, Oslo, Norway, June 17–21, 2000; the 16th annual meeting of the International Society for Traumatic Stress Studies, San Antonio, Tex., Nov. 16–19, 2000; and the Seventh European Conference on Traumatic Stress, Edinburgh, U.K., May 26–29, 2001. Received Dec. 14, 2001; revision received Aug. 19, 2002; accepted Feb. 25, 2003. From the Psychiatric Department and the Department of Psychosocial Medicine, University Hospital, Zurich, Switzerland; and the Division of Neurosciences and Psychological Medicine, Imperial College School of Medicine, West Middlesex Hospital, Isleworth, Middlesex, U.K. Address reprint requests to Dr. Schnyder, Psychiatric Department, University Hospital, Culmannstrasse 8, 8091 Zurich, Switzerland; [email protected] (e-mail). Supported by the Swiss National Science Foundation (project number 32-43640.95). The authors thank Claus Buddeberg, Jasmin Donati, Edgar Heim, Christel Nigg, and Otmar Trentz for their contributions to this study.