OJHAS: 2011-1-5. Latha KS, Ravi Shankar BG. Hospital Related Stress among Patients Admitted to a Psychiatric In-patient Unit in India

Abstract:

The psychiatric patient�s attitudes towards hospitalization have found an association between patient perceptions of the ward atmosphere and dissatisfaction. The aim of the study was to determine the aspects of stress related to hospitalization in inpatients admitted to a psychiatric facility. Fifty in-patients of both sexes admitted consecutively to a psychiatric unit in a General Hospital were asked to rate the importance of, and their satisfaction with, 38 different aspects of in-patient care and treatment. Results showed that the major sources of stress were related to having a violent patient near to his/her bed; being away from family; having to stay in closed wards; having to eat cold and tasteless food; losing income or job due to illness, being hospitalized away from home; not able to understand the jargons used by the clinical staff and not getting medication for sleep. A well-differentiated assessment of stress and satisfaction has implications for the evaluation of the quality of psychiatric care and for the improvement of in-patient psychiatric care.
Key Words: Psychiatric hospitalization; Ward atmosphere; Patient satisfaction

Hospitalization for psychiatric patients is often necessary when it is determined that their behaviors are acutely dangerous to themselves or those in their environment.(1,2) Hospital imposes a special environment in which the meanings of behavior can be easily misunderstood. The consequences to patients hospitalized in such an environment- the powerlessness; depersonalization, segregation, mortification and self-labeling seem undoubtedly counter-therapeutic and stressful.

The stress related to hospitalization may be diverse - due to poor resources (financial or lack of caregivers to stay with the patient) or some variables related to the ward environment. Further, it may be related to having to interact with the various professionals and their inexperience to deal with such situations. These factors singly or in combination may become a hindrance for effective treatment outcomes.

The assessment of patients' satisfaction with medical services has been a rapidly developing area of research for almost 20 years.(3-7) However, the assessment of patients' satisfactionwith psychiatric clinics has only recently started to come into focus.(8-11)

Psychiatric hospitalization is conceptualized as a milieu where there is an experience of trauma .The experience of being a client in a locked inpatient psychiatric unit can be considered traumatic (12,13) which invoke in consumers a response of fear, helplessness, distress, humiliation, or loss of trust in psychiatric staff.

Robins et al (2005) in a semi structured qualitative interviews with 27 randomly selected mental health consumers treated in a psychiatric setting about the descriptions of adverse events experienced while receiving treatment reported two incidents related to the hospital setting including the fear of physical violence and the arbitrary nature of rules. The second set related to interactions with clinical staff including depersonalization, lack of fairness and disrespect.(14)

In a study of the relationship between patients' perception of the real and ideal ward atmosphere and their satisfaction. Patients in locked wards perceived more anger and aggression and patients subjected to coercive measures perceived less autonomy and practical orientation.(15,16)

In a Finnish psychiatric hospital that explored the factors associated with satisfaction patients were quite satisfied with their care. Of seven different satisfaction areas, they were most satisfied with staff�patient relationships, and reported most dissatisfaction in the areas of information, restrictions, compulsory care and ward atmosphere/physical milieu. Younger and female patients were less satisfied with staff�patient relationships than older patients and men.(17)

To the best of our knowledge, there are no studies in our set up related to the factors contributing to stress/ dissatisfaction associated with psychiatric hospitalization. The overall aim of this study was to identify those areas of hospital treatment that patients consider important and/or satisfactory/ unsatisfactory. The understanding of which might help to implement various interventions to tide up with the stress associated with psychiatric hospitalization.

This cross sectional study was conducted at a General Hospital. Kasturba Hospital is a 1470 bed General Hospital serving a Catchment area approximately covering 4 districts. It is attached to a major medical college and is affiliated with the Manipal University, Manipal, India.

The adult inpatient psychiatry unit is housed in the first floor and is independent and a closed ward. The total bed strength is 40 for both sexes and in separate wings. There is also an additional 15 beds within the same ward but separate wing for substance use disorders. In addition, psychiatric patients who are not disturbed and harmful to self or others are admitted to other open medical wards. The bed occupancy is usually full and the average length of stay is 12 days. Most of the patients on an average 95% are discharged into the community and a small number are transferred to long-term facility. The treatment and care of patients are by a multidisciplinary team. The admission process occurs in a comfortable and in a non-threatening atmosphere where the significant others are encouraged to stay with the patients most of the time during their stay. Though there are particular visiting hours, it has been flexible for maximal family involvement. Family psycho education or therapeutic family interactions occur at least twice during their hospitalizations as part of the treatment programmes. It is a closed ward; however there is minimal use of seclusion and restraint.

Subjects for this study (n- 50) were patients admitted to the psychiatric unit of the hospital. Hospitalization is often initiated to maintain patient safety while stabilizing severe adjustment problems. Subjects ranged in age from 16-70 years and a mean of 36.8(�S.D.13.6). The response rate was 89% (50 out of 56).

Of the 50 subjects 29(58%) were males and 21(42%) were females. It is noteworthy that in terms of the primary reasons for admission 26 (52%) were depressed; 10 (20%) were suicidal and 14(28%) displayed significant aggression and were considered to be dangerous to themselves as well as to others. Most of the patients who were included were recovered psychotics or patients with depressive disorders. Severe depressives were excluded for the likelihood of distortions in their responses due to their mood states; neurotic patients who had improved were also taken.

Sociodemographic details such as the age, education, marital status, occupation, family details were gathered with the help of a proforma devised for the purpose in a semi-structured interview.

Hospital Stress Rating Scale (HSRS) (18) developed by Volicer and Bohannon (1975) was used to elicit stressful events related to hospitalization. Holmes and Rahe (1967) (19) used a method from the field of psychophysics to devise. The Social Readjustment Rating Scale, a tool to measure life stress. Volicer and Bohannon (1975) (18) used a procedure comparative to Holmes and Rahe (19) in developing. The Hospital Stress Rating Scale. Several investigators, (20) have used the Volicer (1973) (21) instrument as a patient outcome criterion measure in evaluating nursing practice, thus reaffirming that the techniques used by Holmes and Rahe and Volicer are suitable for use in a clinical setting. These methods were used, therefore, for the development of this instrument. Namely, the utilization of a procedure that assumes persons are able and willing to assign magnitude to, in this instance, stress related to hospitalization. The Hospital Stress Rating Scale has been widely used in other settings too such as patients admitted in medical and surgical wards (22,23); stress in hospitalized AIDS patients.(24)

This is a 49-item scale designed to explore the perception of stressful events related with hospitalization by patients. This scale has 4 sub-scales. There are four main categories such as: 1. Events related to hospitalization- 11 items; Events related to patient-staff relationship-7 items 3. Events related to ward environment- 13 items 4. Events related to special difficulties arising out of the disease- 7 items. Each item was scored on a 3-point scale of severity ranging from Low stress -1; medium stress-2; and high stress-3. The total score was calculated by summing up the scores of all items. The scoring pattern was, higher the total score higher was the stress experienced and lower the total scores lesser was the stress.

After going through the literature and identifying the probable stresses/satisfaction related to psychiatric hospitalization, some of the items were included for example �staying in a closed ward�, �having someone near your bed who is destructive/ violent/ depressed�, �not getting medication for sleep� and �side effects of medications- drowsiness, dryness of mouth, tremors, giddiness and other discomfort�. Some of the items relevant to hospitalization in general were retained. The events related to patients who were serious, medically ill, and not applicable were eliminated and some of the sentences were changed or rephrased to suit the chosen sample. Overall, we had a scale with 38 items, which was thought to be of relevance in our set up.

The scales was translated in the local language by a professional in the field fluent in both languages and later back translated, later verified by another expert. Ambiguous words were changed. All the items were found to be culture free. Of the 49 original items only 38 were included as the rest was not specific to the patients being treated in the psychiatric ward. This again was done in consultation with the experts.

In an attempt to measure patient satisfaction, formal scales though allows for ease ofadministration, contain relatively few items and risks losing potentially important information and may not represent any advantage over asking patients a global question such as �Were you satisfied with treatment?�.(13) Hence, an open-ended question was asked to elicit any other areas with which patient were not satisfied.

Social support assessment: Social support network and other stressors- two questions were related to elicit information about social support network from friends and relatives. The number of relatives and friends with whom they had enduring relationships and were in regular contact were included. There was an open-ended question about stressors-whether the patients experienced any significant stressors in the last 6 months and if the answer was in the positive further details were elicited to know the nature of the stressor.

Assessment of coping strategies was made using a self- report coping scale. This measure was designed to examine the common strategies used based on Roth & Cohen (1986) approach/avoidance model of coping.(25) the approach scale assesses support seeking and problem solving strategies, whereas the avoidance scale reflects emotional distancing and/or attempts to ignore the problem. Eight options were presented and patients who had more than one mode of coping were asked to indicate the same. The number and nature of strategy employed were taken into account.

Institutional Ethical clearance was obtained for the study. Patients and their relatives were explained about the purpose of the study and their willingness to participate was taken. Confidentiality was assured. After their consent was taken, a social worker who was familiar with the protocol met for 30 minutes the patient in the in-patient unit 3-4 days after the hospitalization. During this time, the patient completed paper-and-pencil questionnaires related to hospital-related stressors, coping and social support. Demographic and other clinical information was also collected in a semi-structured interview with the patient and relative.

The data were tabulated and descriptive analysis was carried out. Independent t test and chi square was carried to find the statistical significance across genders on hospital stress scales and coping.

Sociodemographic characteristics: The Sociodemographic characteristics of both the genders are shown in Table 1.

Table 1: Sociodemographic Characteristics of the Sample
Variable	Males(29)		Females(21)		Total(50)
Variable	N	%	N	%	N	%
Religion
Hindu	23	79.3	18	85.7	41	82.00
Christian	4	13.8	3	14.3	7	14.00
Muslim	2	6.9	-	-	2	4.00
Marital status
Single	10	34.5	10	47.6	20	40.00
Married	19	65.5	10	47.6	29	58.00
Widow	-	-	1	4.8	1	2.00
Education
No formal education	1	3.4	2	9.5	3	6.00
Primary	9	31.1	7	33.3	16	32.00
High school	8	27.6	5	23.9	13	26.00
Collegiate	11	37.9	7	33.3	18	36.00
Habitat
Urban	14	48.3	13	61.9	27	54.00
Rural	10	34.5	6	28.6	16	32.00
Suburban	5	17.2	2	9.5	7	14.00

Among both the sexes about 48% were from within the district and the remaining hailed from the neighboring districts and states. The sample comprised of 58% married while 40% were single and the rest widowers. Religion-wise distribution followed the general population � Hindus-82%; Christians- 14%; and Muslims- 4%. The educational backgrounds of the sample were 94% were literate; 6% did not have any formal education; 62% had either high school or collegiate education and one-third primary. 54% resided in urban and the rest from rural or semi-urban localities. The mean age of males were 37.5(S.D. 11.9) and females 35.9 (S.D. 15.8)

Age distribution also revealed that nearly a third of the patients of both genders were in the fourth decade of their life. Majority of males were between 31-50 years � 62.1% compared to females who were either 21-30 -33.4% or 41- > 50 years -38%.

Majority of the males were employed in various institutional set-ups-62.1%; 24% were agriculturists; 6.9% students and 6.9% were unemployed. Among the females 76.2% were homemakers; 4.8% agriculturists and 20% employed in other sectors. There was none who was retired from service.

Among both genders 60% hailed from non-nuclear families, here both extended as well as joint families were clubbed and the rest from nuclear families. In both the sexes the patient him/her self was the head of the family-32%; father in 24%; spouse in 18% and in the rest either the sibling, mother or some one else.

Among the male patients the main bread winner was the patient himself 38%; fathers 24% and other 38% on the other hand among females the breadwinners were either the spouses in the married and fathers in those who were single.

Hospital stress: Table 2 shows the events reported by both the genders, which were stressful related to their hospitalization. In the study patients responded to only 19 items (19 out of 38) as being of high stress, there were no events reported which was of moderate stress and the rest was not stressful. Hence, only these 19 items were taken up for analysis. There was none who reported any events, which were stressful to the open-ended question.

Table 2: Distribution of the Events of Stress in Different Domains
Events of Stress	Males (29)	Females (21)
Events related to hospitalization	No. (%)	No. (%)
Being hospitalized away from home	5 (17.2)	2 (9.5)
Being put in hospital because of illness	4 ((13.8)	-
Not knowing the reasons for treatment	4 (13.8)	-
Not knowing for sure what illness you suffer	4(13.8)	2 (9.5)
Not being told about the diagnosis	4 (13.8)	-
Having to eat cold and tasteless food	3 (10.3)	3 (14.3)
Being hospitalized far away from home	-	3 (14.3)
Patient- staff relationship
Having nurses and doctors use words you do not understand	3 (10.3)	-
Not getting medication for sleep when you need it	3 (10.3)	-
Related to ward environment
Having to stay in closed wards	8 (27.6)	4 (19.0)
Having to stay in bed /ward all the time	7 (24.1)	5 (23.8)
Being aware of the unusual smells around you	7 (24.1)	-
Having a seriously ill patient beside you	6 (20.7)	7 (33.3)
Having a disturbed/violent/depressed pt near you	6 (20.7)	8 (38.1)
Not having enough money to pay bills	6 (20.7)	-
Having a stranger sleep in the same room/ward	3 (10.3)	-
Special difficulties arising out of disease
Thinking about loosing income/job due to illness	8 (27.6)	-
Sudden hospitalization, when not expecting	4 (13.8)	2 (9.5)
Worrying about family members being far away	3 (10.3)	4 (19.0)

Table 3: Distribution According to the areas of Stress - Gender-wise
Areas of Stress	Males	Females	X²	p-value
Event related to hospitalization	24	10	2.000 (DF-1)	.096
Patient-Staff relationships	6	0	2.000 (DF-1)	.096
Ward Environment	43	24	2.000 (DF-1)	.096
Special arising of disease	15	6	2.000 (DF-1)	.096

In the present study on an overall, males reported more number of events in all the 4 domains i.e., 88 events in comparison to females who reported around 40 events which was significant. (F-2.77, df-1, p<.001)

In the subscale i.e., events related to hospitalization males reported events such as being hospitalized far away from home (17.25% ); being put in the hospital because of illness- (13.8%); not knowing the reasons for their treatment (13.8%); not knowing for sure what illness they have- (13.8%); not being told about their diagnosis-(13.8%); having to eat cold or tasteless food- (10.3%).In the females events related to hospitalization such as not knowing when to expect things will be done-(14.3%) and having to eat cold or tasteless food-(!4.3%). The Fisher�s exact test showed that there was no significant difference between males and females on this subscale.

In subscale i.e., events related patient - staff relationship males experienced stress of having nurses or doctors using words that they did not understand - (10.3%); and not getting medication for sleep when patients need it- (10.3%). The females did not perceive any stress in the second sub scale of patient- staff relationship and was not significant.

In the events related to ward environment males reported stress of having to stay in closed wards - (27.6%); having to stay in bed or in the ward all the time- (24.1%); being aware of unusual smells around the patients - (24.1%); having a patient near who is destructive / violent - (20.7%); having beside a seriously ill patient - (20.7%) not having enough money to pay for hospitalization (20.7%). Among females events such as having a patient near you who is destructive /violent / depressed -(19%); having a seriously ill patient beside you -(23.8%); having to stay in the ward all the time -(33.3%); having to stay in closed wards- (38.1%). This subscale was again not significant.

The events related to special difficulties arising of the disease, males expressed their stress of thinking about losing income or job because of illness (27.6%); Sudden hospitalization when they did not expect it (13.8%); worrying about family members being far away from patient (10.3%). The special difficulties arising of the disease experienced among females was related to worrying about family members being away from them in (9.5%); and sudden hospitalization when they were not expecting it (19.%) this was not significant.

Stressors: There was the presence of significant stressors in the last 6 months. Stressors were reported by both the genders i.e. 58%. While among males, 65.5% compared to 47.6% in females. The most common stressor was death of close relative-31.6% in males and 50% in females; losses 52.6% in males and 20% in females.

Social support networks: The findings indicated that a higher proportion of the male patients reported having relatives 6 to more than 10 � 65.6% compared to females where 81.0% had 2-5 persons in their network of social relationships. The network of friends also was higher among males 51.8% had friends about 10, whereas among females the network was 3-5 in 52.4%.

Coping strategies:The Pearson chi square test showed that there was a significant difference between males and females in the nature of coping. For instance males consulted relatives or counselors more often than females and significant at <.05; or resorted to use of substances such as caffeine, nicotine or alcohol(X² .3.60, df-1, p<05). Females indulged in sleeping when faced with stress than males and significant .p <.01 level.

Table 4: Distribution of Nature of Coping Gender-wise
Coping Strategy	Males (N)	Females (N)	X²	p value
Consults relatives/counselors	29	18	4.407 (df-1)	.036*
Think alternatives	3	-	2.31 (df-1)	.128
Worrying	6	10	.415 (df-1)	.520
Be busy with work	3	1	.516 (df-1)	.473
Sleep off	1	6	6.38 (df-1)	.012**
Drink caffeine/ alcohol/ smoking	13	4	3.60 (df-1)	.058*
Eat excessively	-	-		NS
Pray to God	22	17	.184 (df-1)	.666
Some persons reported more than one option; p<.05; *<.01

It was observed that more than one coping strategies were employed by some patients. The finding shows 38% of the patients utilized two strategies while 58% of the patients utilized three strategies to cope with problems.

In the present study, females were younger, tended to be single with less number of years of education in comparison to males. It can be speculated that in women with a rural background, marriage is considered a significant life event that is to be conducted at the appropriate age. Parents consider that it is their commitment to arrange for their daughter�s marriage at all odds and hence might be seeking psychiatric consultation in the early phase of the illness.

The findings that majority of the males were the head of the family and main breadwinner, as most of the families in India is based on patrilineal descent.(26) It is father-centered or father-dominated families.

The study indicated that most of the males reported stress of being hospitalized far away from home and in both sexes stress of worrying about family members. This was supported with previous research which shows that patients encountering unfamiliar environment away from family and admitted against their wishes.(27) In present study about 52% of samples reported that it was inconvenient to stay away from home as most of males were either head of the family and main bread winner. Being hospitalized amounted to loss of income and being from small families the overall management could not be allocated to anyone within the family especially where the children were still young. In addition, it was the hospital policy that a family member stayed with the patient during the hospital tenure.

Some of the males experienced stress like being put in hospital because of illness and sudden hospitalization when they did not expect it. This often happens when hospitalization becomes mandatory for patients who are destructive, violent or uncontrollable and in such instances relatives have to take decisions regarding hospitalization.

Earlier studies found that that psychiatric hospitalization is often necessary when it is determined that their behaviors are acutely dangerous to themselves or for those in their environment.(1,2)

The findings also indicated that the stress for males is associated with not knowing the reasons for their treatment, not being told about their diagnosis. On the other hand, the stress for females was related to not knowing when to expect things will be done. The reasons for this may be two fold as a high proportion of both genders had education up to high school or less. Therefore, they were probably not having enough knowledge about the nature of their illness, its course and outcome. Secondly, they may have been ignorant of the medical procedures and hospital policies. Majority of them hesitated to clarify their doubts with the mental health professionals and distanced themselves from them. As a result, their misconceptions, queries and doubts remained unclarified.

The study has recognized that males and females experienced stress of having to eat cold and tasteless food. This is particularly relevant in view of the patients as they came from far off places and had to be dependent on hospital canteen. Patients came from different places and cultural backgrounds, their selection of diet, cooking styles varied and hence became a source of stress.

Current study also showed that males were being stressed about the patient-staff relationship where they expressed that they were bothered about the nurses and doctors using words that they could not understand. This may be the effects of the language barriers, differences in dialects, accent, and usage of technical words by the professionals. This leaves the patients and their relatives ignorant and confused about various illness- related information as there is no proper communication between the staff and the patients. As noted by Cleary and McNeil (1988), (28) higher satisfaction may be a result of better patient-physician interactions.

Males reported stressed about not getting medication for sleep when they needed it. In most psychiatric illnesses, sleep disturbances are the most common symptom. The prescription for sedative is suggested by doctors and not given on the demand of the patients. In such situations due to sleep disturbances, which might be mild or severe, patients complain of feeling tired, dull, the next morning, which may be present throughout the day. It is a popular belief that a good night�s sleep is basic of the needs, which rejuvenate human beings.

In the domain of ward environment both the genders reported stress of having to stay in bed and in closed wards all the time. This study was conducted in the psychiatric inpatients, which is a closed ward set-up. Here a variety of patients are admitted with a variety of psychiatric problems- such as suicidal, harming others, destructive, violent, or trying to escape from the ward and for these reasons hospitalization is initiated to maintain patient�s safety while stabilizing severe adjustment problems. In such cases, patients are kept in closed wards under maximum staff observation. In an earlier study by Causey (1998) (29) showed that stress in psychiatric patients in a psychiatric setting was because of being in a place where all the doors were locked. In another study by Drake & Wallach (1988) (30) reflected that the ward atmosphere in the sample he studied was considered stressful. The other reasons could be sudden withdrawal from the routine housework for females and absence of outside employment for males due to which hospitalization was perceived as stressful.

In the sample studied both the sexes reported stress about having a seriously ill patient and having a patient near their bed who is destructive or violent. It is but normal to be fearful as it is observed that psychiatric inpatients often display significant degree of aggression, violent behaviours, dangerous or threatening to others, or making loud noises. They also tend to be impulsive and unpredictable in their behaviours. All these seems stressful to the patients who are hospitalized.

It was an observation in this study that males reported stress because of not having enough money to pay for hospitalization. The result shows that in more than fifty percent of the patients examined the monthly income was between 1000-3000 rupees. This would be a meager sum and it would be difficult to meet all the expenses of the family with the cost of all essential commodities increasing. As in most instances, the head of the family was the sole earner. In such cases patients are left with no savings and have to incur loans to meet the hospital expenses and other emergencies. On the other hand, females reported that their areas of stress were sudden hospitalization when they were not expecting it. It is likely that some of the patients had come for a follow-up on aggravation of their symptoms and on advice of the psychiatrists get hospitalized immediately for which they are not mentally prepared. Unplanned hospitalization results in being admitted without essential things like clothes, toiletry or finances. The unfamiliar ward atmosphere and staff interactions add to the stress. This is especially true in case of patients who have been admitted for first time.

Studies suggest that the factors found to contribute to satisfaction include ward atmosphere, certain Sociodemographic characteristics, diagnosis, duration of illness, and previous in-patient treatment.(31-34)

The results show that both genders tried coping with the problem by consulting with relatives, friends and counselors. It was also evident among males that more than fifty percent reported of having more than 10 persons in their network of friends and 93% reported having similar network size of relatives too, on the other hand among females about 52% reported of having 3-5 persons in their network of friends and 90.5% reported having similar network size of relatives. Usually the family and family members are the most important primary group and persons try to confide with them in periods of crisis. However, this again is dependent on their closeness with their kith and kin. In a closely-knit and cohesive family unit, this might be possible. Nevertheless, there may be many instances where professional help is often very essential to person cope with a problem.

The findings of the study also indicated that majority of males and females were praying to God to cope with their problems. It is a general observation that in times of crisis many resorts to religious coping and it have become ingrained in our upbringing and culture too. In almost all religions saying daily prayers, performing rituals are regularly practiced. It was also seen in this study that males consumed excessive coffee, tea or alcohol. It can further be thought that most of them were not educated enough to know about the ill effects of all these beverages and some of them had misconceptions that it reduces internal tensions and induces sleep. Many of them were ignorant of the long term effects of consumption of these beverages.

This study also pointed that females coped with problem by worrying and not trying to find solutions. This probably could be explained as the females being home- makers had no resources to lean upon and hence no control over the situation. Some of them confined to the house. So there was no chance for them to mingle with others or share their feelings.

The present study enabled to gain an understanding of the stress related to hospitalization among psychiatric in-patients, and their coping patterns and social support networks. The finding of the study has implications for mental health professionals.

Professionals have to inform the patient and their relatives the reasons for their hospitalization, referral to specialized treatment, advantages of hospitalization, approximate plan of admission and discharge, minimum expenses of the treatment, hospital procedures and other special benefits from the hospital which would probably predict better treatment outcomes. Developing a plan of care, assessment of stresses in hospitalization and understanding patient�s background may be helpful to deal with hospital stress in patients.

Patients and staff relationship is based on good rapport, proper communication and language. The professionals should explain about the illness and treatment issues in simple words and in a language familiar to the patient. Group meetings between patients, family members and the staff would help to sort the difficulties related of hospitalization.

In-patients' attitudes towards their psychiatriccare involves a complex relationship between clinical and sociocultural characteristics. In view of the multifarious problems encountered by the hospitalized psychiatric patients, it is imperative on the part of the multidisciplinary psychiatric team, especially the psychiatric professionals to have a clear understanding of the various problems and stresses confronting them. It thus becomes necessary to prepare patients for the event i.e., hospitalization.