INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Asthma is a chronic respiratory disease characterized by airway obstruction, airway inflammation, and bronchial hyperresponsiveness, and can vary from a very mild disorder to a disabling and life-threatening disease. Only a small minority of patients with asthma has severe disease, which nevertheless has a profound impact on health status (1) and accounts for more than 50% of direct and indirect costs of asthma (2, 3). The high costs of severe asthma appear to be related to the loss of asthma control (emergency room visits, hospitalizations) rather than to the management of controlled asthma (drugs, GP care) (2, 4). It therefore seems important to better define the characteristics of these uncontrolled high-cost patients with asthma.

Several psychosocial and emotional factors have been noted to be associated with poor asthma control, near-fatal asthma attacks, and asthma mortality. These factors include denial (5), anxiety (6), hypochondriasis (7) and inappropriate coping skills (8), probably leading to noncompliance with medications, irregular follow-up, and suboptimal asthma management.

The clinically relevant question remains whether or not psychopathology in patients with severe asthma is a risk factor for loss of asthma control with subsequent increased health care utilization. Therefore, we investigated health care utilization in a large group of outpatients with severe asthma, and compared patients who fulfilled the criteria for psychiatric caseness according to the short form General Health Questionnaire with patients who did not.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Patients with severe bronchial asthma as defined by the ERS Task Force on difficult/therapy-resistant asthma (9), aged 18-75 yr, were asked to participate in this study. They were recruited from the outpatient pulmonary departments of 10 hospitals (2 teaching and 8 nonteaching) in the western part of The Netherlands between May 1998 and June 1999 and were a representative sample of the Dutch population, mixed rural as well as urban citizens. Patients were allowed to enter the study if they had a diagnosis of asthma, based on a history of episodic dyspnea and wheezing, hyperresponsiveness to inhaled histamine (provocative concentration causing a 20% fall in FEV₁ [PC₂₀] < 8 mg/ml), and/or a documented bronchodilator reversibility in forced expiratory volume in 1 s (FEV₁) of > 12% predicted (10). All patients had to use high doses of inhaled corticosteroids (dose  1,600 µg/d beclomethasone or equivalent) and long-acting bronchodilators for more than 1 yr. Despite this treatment, they had to be symptomatic with an impaired exercise tolerance, nocturnal symptoms, and/or frequent use of rescue medication. Furthermore, they had to have difficult-to-control asthma, with at least one severe exacerbation requiring a course of high dose oral corticosteroids during the previous 12 mo or maintenance therapy with oral prednisone  5 mg/d. Current smokers and patients with a smoking history of more than 10 pack-years were excluded from participation, as were those patients (n = 7) who were not fluent in Dutch. The study was approved by the Hospital Medical Ethics Committee, and all patients gave informed consent.

Design

For the present study the patients visited our hospital during a whole day as part of an extensive study on severe asthma, aimed at identifying risk factors for asthma severity and persistent airflow limitation in asthma.

First, the characteristics of the patients and their asthma were documented according to a structured questionnaire always taken by the same investigator. Then the patients completed self-report questionnaires on psychological functioning and health care utilization. Finally, a blood sample was taken, and lung function measurements were performed.

Psychological and Health Care Utilization Questionnaires

The General Health Questionnaire (GHQ) is the most widely used screening test to detect psychiatric disorder in medical practice and measures possible prevalence of nonpsychotic psychiatric disturbances, especially anxiety and depression disorders. We made use of the short GHQ-12 version (11) in which each of the 12 items (Table 1) with 4 answering categories was scored on a bimodal response scale, resulting in a score ranging from 0 to 12 (12). A cut-off score of 5-6 was used to discriminate possible psychiatric cases (score  6) and noncases (score  5), which according to a GHQ evaluating study shows a positive and negative predictive value of 71% and 78%, respectively (13).

The health care utilization questionnaire asked for data concerning the last year on hospital admissions, use of (chronic or courses of) oral corticosteroids, and frequencies of asthma-related visits to a general practitioner (GP), a chest physician, and an emergency room. Additionally, data about intensive care unit (ICU) admission or mechanical ventilation in the past were collected.

Atopy

Atopic status was assessed by specific immunoglobulin E (IgE) to a panel (house dust mite, grass and birch pollen, cat and dog dander, mugworth and fungi) of common aeroallergens (Phadiatop; Pharmacia, Uppsala, Sweden).

Pulmonary Function

Bronchodilators were withheld, if possible, for at least 12 h before pulmonary function tests were performed. Slow inspiratory vital capacity (VC) and FEV₁ were measured by a dry rolling seal spirometer. FEV₁/ VC was assessed 30 min after the administration of 400 µg salbutamol and 80 µg ipratropium bromide using a metered-dose inhaler with a spacer device. The predicted values for all parameters were calculated according to the proposal of Quanjer and coworkers (10).

Statistical Analysis

Differences between psychiatric cases and noncases were analyzed using unpaired Student's t tests, chi-square analyses and nonparametric tests, when appropriate.

The following contrasts in health care utilization in the last year were considered for psychiatric cases versus nonpsychiatric cases: four or more visits to a GP versus fewer than four visits to a GP, four or more visits to a chest physician versus fewer than four visits to a chest physician, two or more emergency visits versus none or one emergency visit, two or more exacerbations versus one exacerbation, two or more hospital admissions versus no or one hospital admission, maintenance therapy with oral corticosteroids versus no chronic oral steroids, ever ICU admission versus never ICU admission, and ever mechanical ventilation versus never mechanical ventilation. Odds ratios for psychiatric cases versus nonpsychiatric cases as reference group were obtained by multiple logistic regression analyses with one of the health care utilization parameters as dependent at a time in the model. To correct for possible confounding and/or effect modification by age, sex, and FEV₁, these factors were added in the model to obtain odds ratios adjusted for these factors. All analyses were performed using the Statistical Package of the Social Sciences (SPSS-9.0). p Values less than 0.05 were considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

In the whole group of 98 patients with severe asthma the median (range) score on the GHQ was 2.0 (0-12). According to the GHQ cut-off score 21 (21.4%, 95% CI: 14%-31%) of the 98 patients were considered as psychiatric cases, with a median (range) GHQ score of 8 (6), whereas 77 (78.6%) patients showed no psychiatric disorder, with a median GHQ score of 0 (0-5). Table 2 demonstrates the characteristics of the psychiatric cases and noncases within the patients with severe asthma who participated in this study. Recruitment was equal from teaching as compared with nonteaching hospitals, and there was a wide range in age (18-75 yr), age at onset of asthma (0.5- 68 yr), and asthma duration (2-67 yr) in the total group. Overall, there were no significant differences between psychiatric cases and noncases in demographic and objective disease characteristics. Although the predominance of females was higher in the psychiatric patients with severe asthma (86%) as compared with the nonpsychiatric control patients (68%), this difference did not reach the significance level.

Differences in Health Care Utilization

Seventy-one percent of the psychiatric cases visited their GP four or more times during the last year as compared with 30% of the nonpsychiatric patients with severe asthma (crude OR = 5.9) (Tables 3 and 4). In addition, 71% of the cases needed two or more emergency visits versus 31% of the noncases (OR = 5.3). Almost all psychiatric patients with severe asthma (92%) had two or more asthma exacerbations in the previous year, as compared to 57% of the nonpsychiatric patients (OR = 12.4), whereas no significant difference between the groups was found for the use of maintenance oral corticosteroids (38% versus 34%, OR = 1.4). Finally, 19% of the cases versus 5% of the noncases needed two or more hospital admissions in the past year (OR = 4.8).

After adjustment for age, sex, and FEV₁, it appeared that psychiatric patients in comparison with the nonpsychiatric patients with severe asthma had an increased odds ratio for four or more asthma-related visits to their GP and for two or more emergency visit of 6.7 and 4.7, respectively (Table 4). The psychiatric cases among the patients with severe asthma had an almost 11 times increased risk for two or more asthma exacerbations and an almost 5 times increased risk for two or more hospitalizations during the past year as compared with the patients with severe asthma without psychiatric disorder. Furthermore, we observed an adjusted odds ratio of 16 in the psychiatric cases for a positive history of mechanical ventilation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results of the present study show an evident association between psychological dysfunctioning and health care utilization within a large, well-defined group of outpatients with severe asthma. The psychiatric cases as compared with the nonpsychiatric patients with severe asthma had considerably more visits to a GP, emergency visits, prednisone courses for asthma exacerbations, and hospitalizations, including ICU admissions and mechanical ventilation, all known determinants of direct costs of asthma (2). The present findings suggest that the morbidity and costs of asthma might be related to the level of psychological dysfunctioning in patients with severe asthma rather than to asthma severity per se, thereby identifying an area of potential intervention.

To our knowledge, this is the first study investigating the contribution of psychopathology to health care utilization within a regularly supervised group of patients with severe asthma. Until now, the relationship between psychological disturbances, asthma severity, and some measures of health care utilization has been examined only in patients with near fatal or brittle asthma (5, 8, 14, 15). Also in our study patients with severe asthma with psychiatric disorder had more near fatal attacks in the past. Our results show that the unfavorable effect of concurrent psychopathology on health care utilization is relevant to a more general group of patients with severe asthma, namely patients who repeatedly visit their chest physicians in general hospitals, thereby representing the population of patients with severe asthma that is in reach of chest physicians and may be accessible for intervention to reduce morbidity and costs of asthma.

The results of the present study do not seem to be biased by the methods we used. The GHQ is a well-validated questionnaire. Although it cannot differentiate between specific types of psychiatric disorders, the GHQ is very useful in detecting possible psychiatric cases in medical practice, and has good positive and negative predictive values (13). We observed a prevalence of psychiatric cases of 21% in our general population of patients with severe asthma, which is comparable to the prevalence found in the normal Dutch population (20%) (16) and in internal medicine outpatients (15%) (17), suggesting that asthma in itself is not associated with increased psychiatric disorder. The higher prevalences found in studies on brittle (40%) (15) or near fatal asthma (43%) (14) seem to be associated with patients with extremely unstable or severe asthma only, and can be confirmed in our study as we found 36.4% of psychiatric cases in the subgroup of patients with at least three exacerbations in the previous year (data not shown). Yet, the prevalence of psychological disturbances we found might have been negatively influenced by the exclusion of current smokers and patients with a smoking history of more than 10 pack-years, thereby possibly excluding psychiatric cases as psychiatric disorders are associated with a high rate of cigarette smoking (18).

The data on health care utilization are based on self-report, and therefore may be influenced by recall bias. However, previous research has shown that recall of health care utilization data in respiratory patients is fairly reliable for hospital admissions and visits to chest physicians but to some lesser extent for visits to emergency room and GP (19), whereas for the latter a bias toward underreporting at higher numbers of visits has been suggested (20). So, regarding the association of psychopathology and health care utilization in severe asthma, both the exclusion of smokers and the possible recall bias are only expected to lead to underestimation of the dimensions of the problem.

We demonstrated an evident association between psychopathology and increased health care utilization in severe asthma, but as this study was cross-sectional in design we can only speculate about the mechanisms. More longitudinal studies are required to clarify the causality of this association.

An independent, direct relationship between health care utilization and psychopathology could be postulated as similar associations were observed in other chronic diseases (21). However, one could also consider disease-specific relationships, in this case between asthma severity and psychopathology. First, an explanation by a biological link between stress, emotions, and asthma can be suggested. For example, stressful life events have been shown to alter the pattern of cytokine responses in patients with asthma, with a shift toward a Th2-like pattern (22). However, our understanding of the effect of psychological stressors on neural and inflammatory processes is not yet sufficient (23). Second, an effect of psychological factors on symptom perception in asthma is demonstrated by some authors (26), but contradicted by others (27). So, its relevance and potential mechanism remain unclarified. Finally, denial, panic-fear, depression, and inappropriate coping skills are specific psychological features related to persistently suboptimal asthma management (8, 14, 28). This is probably most clearly shown in the association between psychological factors and poor compliance with treatment in asthma (29), which will most likely contribute to poor control of the disease and subsequently increased use of health facilities.

On the other hand, instead of being the cause, psychological disturbances might result from the loss of control in asthma. For example, the high level of denial and anxiety reported in patients following a life-threatening attack might be related to posttraumatic stress reactions (5). Furthermore, asthma medication might have an adverse effect on psychological functioning. The literature contains very little information on such effects in adults, but studies on asthma medication in children indicate medication-specific effects in some aspects of neuropsychological functioning (30). Therefore, such effects in adults cannot be excluded.

The challenging question behind this study was to better define the characteristics of the patients with asthma with high health care utilization. Our results provide evidence that psychological dysfunctioning is one of these characteristics, which, independent of asthma severity, contributes to increased morbidity and costs of asthma. Similar findings have been observed in other chronic diseases (21). The relative factors contributing to the interaction between psychopathology and health care utilization cannot be deduced from the GHQ. More studies using quality of life questionnaires and questionnaires that score anxiety and depression are needed in this group of patients with asthma to further analyze the importance of different contributing factors. If psychological disturbances are recognized and treated in these patients, the morbidity and thereby the costs of asthma might be significantly reduced (31). Therefore, it might be worthwhile to screen for psychopathology in the general work-up for severe, difficult-to-control asthma to increase identification of treatable disorders.

In conclusion, we have shown an evident association between psychological dysfunctioning and increased health care utilization within a large group of patients with severe asthma. This finding concerns the patients with severe asthma in general, visiting both teaching and nonteaching hospitals, and being accessible for potential intervention. Therefore, prospective studies to investigate the value of screening and the effects of treatment of psychological disturbances in patients with severe asthma are necessary.