Health Care Consumption, Psychiatric Diagnoses, and Pharmacotherapy 1 and 2 Years Before and After Newly Diagnosed HIV: A Case-Control Study Nested in The Greater Stockholm HIV Cohort Study

INTRODUCTION

Psychiatric illness has been proposed to be a risk factor for acquiring human immunodeficiency virus (HIV) (1–4), and persons admitted for severe psychiatric illness have been shown to have a higher HIV prevalence than the general population (5,6). Depression is of particular importance in people living with HIV (PLWH), and the prevalence of a diagnosis of depression in PLWH is three times higher in men (10%) and 40% higher in women (8.5%) than in the general population of Stockholm (7). Depending on methodology used and population studied in the world, the prevalence of depressive symptoms in PLWH has been shown to range from 12.8% to 78%, and in a review, no significant differences were noted by country income group (8).

Mental health is of importance in PLWH. Concomitant psychiatric illness may contribute to poor adherence to medication and safe sex practices in PLWH (4,9–11). Both adherence to combination antiretroviral therapy (ART) and adherence to psychotropic drugs are affected by the mental health. In fact, psychiatric illness may affect PLWH in all stages of disease, including disclosure of HIV diagnosis through the whole process related to ART, readiness to start treatment, and lifelong adherence to medication (12,13). Psychological distress after HIV diagnosis is common and has been associated with the stigma that still exists toward PLWH and worries for HIV complications (1,3,4,14). The stigma in PLWH has been shown to be associated with anxiety, depression, and suicide ideation in studies from all over the world (15). In addition, there is a biological explanation to the higher levels of depression, cognitive effects, and fatigue in PLWH, as PHLW have been shown to have impaired neurotransmitter biosynthesis, especially tryptophan and phenylalanine (16).

We reported in previous studies of “the Greater Stockholm HIV cohort,” a cohort that includes all PLWH in Stockholm Region, that, although the prevalence of diabetes, hypertension, and cancer in PLWH is on a par with that of the general population, PLWH more often have psychiatric disorders (7,17).

By matching cases with newly diagnosed HIV to age-matched controls without HIV, we have the opportunity to conduct a case-control study and see how their health care compares in the years before and after the first recorded HIV diagnosis. Do PLWH have psychiatric diagnoses before their HIV diagnosis, or do psychiatric diagnoses follow and consequently occur after the HIV diagnosis?

Accordingly, the aim of this investigation was to study the frequency of health care visits, the types of clinics visited, relevant psychiatric comorbidities, and prescribed and collected psychiatric pharmacotherapies in all individuals with an incident HIV diagnosis compared with matched HIV-negative controls from 2011 to 2018. We hypothesize that there will be differences between PLWH and controls before the time point of HIV diagnosis and that the differences in psychiatric illness between cases and controls will increase after the diagnosis of HIV.

METHODS

“The Greater Stockholm HIV Cohort Study” is an initiative to provide longitudinal information regarding the health of PLWH in comparison to the health of the total population. The Stockholm Region has more than 2.2 million inhabitants, representing more than one-fifth of Sweden’s entire population. The region includes the capital city of Stockholm and several other cities and towns, as well as large rural areas and a sparsely populated archipelago. The Stockholm Region is responsible for financing primary and secondary health care, mainly through taxes. Apart from very few private clinics that operate without subsidies in Stockholm, all consultations and diagnoses are recorded and stored in a central regional database, the Stockholm Regional Health Care Data Warehouse (VAL). The link to VAL makes it possible to perform prevalence and incidence studies for different diagnoses for all residents. These databases compile and store data on health care utilization from primary care, specialist open care, hospital inpatient care, and data on collected prescribed medications. As an indication for its accuracy and validity, VAL is used by the Stockholm Region for updating the National Patient Register kept by the Swedish National Board of Health and Welfare, as well as the annual benchmarking reports of the Swedish National Board of Health and Welfare and the Swedish Association of Local Authorities and Regions. Since 1997, diagnoses have been coded according to World Health Organization’s International Classification of Diseases, Tenth Edition.

Study Population

The studied cohort in the present study was defined as all men and women who resided in the Stockholm Region at some point between January 1, 2011, and December 31, 2018, and subjects were included with their controls when a first diagnosis of HIV was registered during that time frame. Data on all health care consultations in primary care and specialized open care were extracted from the VAL up to 2 years before and after the first HIV diagnosis was registered. Men and women with a newly diagnosed HIV were used as cases. Approximately 10 controls per case, matched for sex, age, and socioeconomic status without e recorded diagnosis of HIV, were used. Each control was only enrolled once, even if matched with more than one of the cases. Therefore, the number of controls is less than 10 times the number of cases.

Sociodemography

We used the Mosaic tool to classify neighborhood socioeconomic status into three levels, that is, high, middle, or low. Mosaic is a tool developed by the marketing company Experian to classify consumers to make sale activities more effective. The mosaic system makes it possible to achieve a nuanced classification of socioeconomic status. It uses multivariate modeling, using more than 400 variables, to group postcodes into different types and aggregate broader groups. Mosaic uses data from 29 different countries and has been shown to be useful for the classification of cohorts in epidemiologic research (18,19).

Design

This was a case-control study in which individuals with newly diagnosed HIV were compared with controls matched for sex, age, and neighborhood socioeconomic status without such a diagnosis. Diagnoses were registered at discharge from hospital or after a consultation 1 year (1–365 days), referred to as year 1, and during 1 year before 1 year before (366–730 days), referred to as year 2, before the first diagnosis of HIV was recorded. We also studied registered diagnoses with the same variables 1 and 2 years after HIV diagnosis. The following International Classification of Diseases codes were used to define the individuals diagnosed with HIV: B20, B21, B22, and B24.

Visits to Health Care and Clinics

The number of visits to doctors was analyzed for the following types of clinics: emergency care, psychiatric care, gynecology (women), other open care (nonprimary care doctors outside hospitals), and primary care. We had no clear information on specific visits to clinics for PLWH only; they are included in the other open care visits. We also analyzed the total number of visits to health care subdivided into physicians; psychologists, therapists, and social workers; and other health care workers (nurses, physiotherapists, and more).

Psychiatric Diagnoses and Pharmacotherapy

To portray the differences in psychiatric illnesses that could be problematic in PLWH between cases and controls, the following comorbid diagnoses were chosen (20): reactions to stress; anxiety disorders (phobic anxiety and generalized anxiety), psychotic disorders (schizophrenia, acute and transient psychotic disorders, schizoaffective disorder, other psychotic disorder not due to a substance or known physiological condition, and unspecified psychosis not due to a substance or known physiological condition), bipolar disorders (mania without psychotic symptoms and bipolar affective disorders), alcohol use disorder, substance use disorders (opioids, cannabis, sedatives, cocaine, other stimulants, hallucinogens, nicotine, inhalants, and multiple drug use and use of other psychoactive substances), depression (depressive episode and major depressive disorder), and self-harm behaviors.

ATC codes were used to identify pharmacotherapy for sleep disturbances and psychiatric disorders: first-generation sedating anti-histamines (R06AD02, N05BB), neuroleptics (N05A), propiomazine (N05CM06)(sleep medication), antidepressant drugs (N06A), hypnotics (N05CF), stimulants(N06BA), and benzodiazepines (N05BA).

Ethics

All data were pseudonymized, and none of the individuals could be identified. Patient consent was nonapplicable because it is mandatory to be included in the registers when in contact with public health care in Sweden. Management and analysis based on VAL is part of a continuous quality control of health care utilization in the Stockholm Region. Ethical approval was obtained from the regional ethical review board in Stockholm to study diseases and their comorbidities with VAL data (permits: 2013/2196-31/2, 2016/638-32).

Statistical Methods

Means and 99% confidence intervals (CIs) were estimated to compare the number of health care visits with all professionals among individuals newly diagnosed with HIV compared with HIV-negative controls. No overlap in the confidence intervals between cases and their controls was considered statistically significant. Conditional logistic regression was used to calculate odds ratios (ORs) with 99% CIs of psychiatric comorbidities and drug prescriptions among individuals newly diagnosed with HIV compared with HIV-negative controls, separated by sex. Multiplicative interactions were tested for possible interactions between sex and diagnoses before and after newly diagnosed HIV. To reduce the risk of false discoveries due to multiple testing, a p value of less than .01 was considered as significant. All controls were matched for age, sex, and neighborhood socioeconomic status. Statistical analysis and data management were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, North Carolina).

RESULTS

In total, 450 women and 930 men were diagnosed with HIV for the first time during the study period and were matched to 4361 women and 9004 men without a diagnosis of HIV. The age distribution, neighborhood socioeconomic status, and prevalence and pharmacotherapy of psychiatric disorders are shown for men and women (cases and controls) in Table 1. The similarity in characteristics that we matched for indicates that the matching procedure was well performed. Substance use disorder was more common in cases than controls among both women and men before HIV diagnosis; other noteworthy differences were not found.

TABLE 1 - Baseline Characteristics in Women and Men (Cases and Controls) 1 Year Before HIV Diagnosis Women Men Frequency 1 y Before Diagnosis for Cases (n = 450) Frequency 1 y Before Diagnosis for Controls (n = 4361) Frequency 1 y Before Diagnosis for Cases (n = 930) Frequency 1 y Before Diagnosis for Controls (n = 9004) Age (mean and standard deviation) 37.17 (12.19) 37.17 (12.29) 39.85 (13.44) 39.93 (13.41) Neighborhood socioeconomic status, n (%)  High 110 (24.4) 1069 (24.5) 374 (40.2) 3618 (40.2)  Medium 82 (18.2) 791 (18.1) 173 (18.6) 1651 (18.3)  Low 258 (57.3) 2501 (57.4) 383 (41.2) 3735 (41.5) Diagnoses, n (%)  Suicide attempt (X60–84) 2 (0.44) 7 (0.16) 1 (0.11) 3 (0.03)  Substance use disorder (F11–F19) 8 (1.78) 13 (0.30) 16 (1.72) 36 (0.40)  Alcohol use disorder (F10) 6 (1.33) 17 (0.39) 17 (1.83) 100 (1.11)  Depression (F32, F33) 20 (4.44) 224 (5.14) 24 (2.58) 214 (2.38)  Fatigue syndrome (F43.8) 2 (0.44) 51 (1.17) 4 (0.43) 26 (0.29)  Stress (F43) 10 (2.22) 153 (3.51) 6 (0.65) 119 (1.32)  Anxiety (F40, F41) 15 (3.33) 238 (5.46) 27 (2.90) 252 (2.80)  Bipolar disorder (F30, F31) 2 (0.44) 25 (0.57) 2 (0.22) 46 (0.51)  Psychosis (F20, F23, F25, F28, F29) 2 (0.44) 24 (0.55) 5 (0.54) 62 (0.69) Pharmaceutical drug type, n (%)  Stimulants (N06BA) 5 (1.11) 51 (1.17) 9 (0.97) 93 (1.03)  Neuroleptics (N05A) 15 (3.33) 89 (2.04) 12 (1.29) 163 (1.81)  Benzodiazepines (N05BA) 13 (2.89) 129 (2.96) 22 (2.37) 179 (1.99)  Tranquilizers (R06AD02, N05BB) 28 (6.22) 256 (6.87) 29 (3.12) 278 (3.09)  Hypnotics (N05CF) 18 (4.00) 238 (5.46) 68 (7.31) 346 (3.84)  Propiomazine (N05CM06) 11 (2.44) 108 (2.48) 24 (2.58) 171 (1.90)  Antidepressant drugs (N06A) 34 (7.56) 483 (11.08) 49 (5.27) 534 (5.93)

HIV = human immunodeficiency virus.

Interactions for between sex and different diagnoses before and after new HIV diagnosis was tested and revealed significant interactions between sex and depression, 1 and 2 years after HIV diagnosis, p < .001 and p = .004, respectively. Significant interactions between sex and diagnoses were also found for anxiety and stress 1 year after diagnosis, p = .009 and p = .003, respectively. All results hereinafter were stratified by sex.

The number of health care visits, types of clinics visited, and the category of health care professionals met on their visits 2 years and 1 year before the first HIV diagnosis and 1 and 2 years after the first HIV diagnosis are shown in Table 2 (women) and Table 3 (men). The results are summarized hereinafter.

TABLE 2 - Number of Health Care Visits and Health Care Professionals Visited With 99% CIs (Women) Clinics Visited and Professionals Met on the Visit 2 y Before Diagnosis (n = 450), Cases (99% CI) 2 y Before Diagnosis (n = 4361), Controls (99% CI) 1 y Before Diagnosis (n = 450), Cases (99% CI) 1 y Before Diagnosis (n = 4361), Controls (99% CI) 1 y After Diagnosis (n = 450), Cases (99% CI) 1 y After Diagnosis (n = 4361), Controls (99% CI) 2 y After Diagnosis (n = 450), Cases (99% CI) 2 y After Diagnosis (n = 4361), Controls (99% CI) Clinics visited  Emergency clinic visits 0.53 (0.24–0.82) 0.47 (0.42–0.52) 0.85 (0.54–1.16) 0.50 (0.45–0.54) 1.36 (0.07–2.65) 0.42 (0.38–0.47) 0.64 (0.38–0.90) 0.35 (0.31–0.38)  Outdoor clinic visits 5.79 (2.67–8.91) 7.47 (6.93–8.01) 7.86 (4.52–11.21) 7.72 (7.15–8.28) 21.73 (17.79–25.68) 7.18 (6.64–7.72) 12.06 (8.06–16.07) 5.78 (5.34–6.22)  Psychiatry clinic 2.87 (0.03–5.77) 1.57 (1.22–1.91) 3.53 (0.46–6.59) 1.56 (1.21–1.90) 4.61 (0.60–8.62) 1.49 (1.13–1.85) 3.58 (0.08–7.08) 1.06 (0.82–1.29)  Gynecological clinic 0.25 (0.13–0.37) 0.50 (0.45–0.56) 0.45 (0.26–0.62) 0.58 (0.52–0.65) 1.00 (0.80–1.20) 0.50 (0.45–0.56) 0.78 (0.62–0.93) 0.42 (0.37–0.47)  Other specialist clinic 1.87 (1.14–2.60) 3.32 (3.01–3.63) 2.88 (2.21–3.54) 3.43 (3.11–4.27) 15.60 (13.44–17.37) 3.24 (2.95–3.53) 7.00 (5.37–8.63) 2.65 (2.38–2.91)  Primary health care 1.32 (0.80–1.85) 2.55 (2.38–2.72) 1.86 (1.44–2.29) 2.64 (2.46–2.82) 1.88 (1.33–2.42) 2.37 (2.21–2.54) 1.35 (0.92–1.79) 2.00 (1.84–2.17) Professionals met on visits  Psychologist, therapist, social worker 0.54 (0.12–0.97) 0.79 (0.63–0.95) 0.68 (0.25–1.10) 0.86 (0.70–1.03) 3.44 (2.72–4.16) 0.78 (0.64–0.92) 1.33 (0.75–1.91) 0.63 (0.51–0.75)  Physician 2.67 (1.87–3.48) 3.88 (3.64–4.11) 4.04 (3.09–4.99) 4.05 (3.82–4.23) 9.04 (7.77–10.30) 3.67 (3.45–3.88) 4.70 (4.02–5.39) 3.01 (2.82–3.20)  Other 3.10 (0.51–5.69) 3.27 (2.94–3.61) 4.00 (1.18–6.82) 3.30 (2.94–3.65) 10.61 (7.04–14.18) 3.16 (2.80–3.53) 6.67 (3.25–10.09) 2.50 (2.23–2.76)

CI = confidence interval; HIV = human immunodeficiency virus.

Comparisons in all women with new HIV diagnosis compared with matched HIV-negative controls 2 years (366–730 days) and 1 year (1–365 days) before diagnosis and 1 year (1–365 days) and 2 years (366–730 days) after the first registered HIV diagnosis, during the period 2011–2018. Significant differences were defined as no overlap in the 99% CIs between cases and controls at each time point and are shown in bold.

No statistical tests were made. Nonoverlapping 99% CIs were considered significant.


TABLE 3 - Number of Health Care Visits and Health Care Professionals Visited With 99% CIs (Men) Clinics Visited and
Professionals Met on the Visit 2 y Before Diagnosis (n = 930), Cases (99% CI) 2 y Before Diagnosis (n = 9004), Controls (99% CI) 1 y Before Diagnosis (n = 930), Cases (99% CI) 1 y Before Diagnosis (n = 9004), Controls (99% CI) 1 y After Diagnosis (n = 930), Cases (99% CI) 1 y After Diagnosis (n = 9004), Controls (99% CI) 2 y After Diagnosis (n = 930), Cases (99% CI) 2 y After Diagnosis (n = 9004), Controls (99% CI) Clinics visited  Emergency clinic visits 0.27 (0.20–0.34) 0.34 (0.31–0.38) 0.71 (0.60–0.82) 0.35 (0.31–0.38) 0.56 (0.43–0.70) 0.31 (0.28–0.34) 0.40 (0.31–0.50) 0.26 (0.23–0.28)  Outdoor clinic visits 5.70 (3.53–7.88) 4.14 (3.81–4.47) 8.35 (6.18–10.52) 4.32 (3.99–4.65) 19.99 (17.16–22.82) 3.93 (3.62–4.24) 10.64 (8.39–12.90) 3.30 (3.06–3.54)  Psychiatry clinic 2.74 (0.76–4.73) 1.10 (0.87–1.33) 3.16 (1.18–5.15) 1.10 (0.86–1.34) 4.23 (1.71–6.89) 1.03 (0.80–1.27) 2.98 (1.06–4.86) 0.74 (0.60–0.88)  Other specialist clinic 2.03 (1.36–2.92) 1.58 (1.50–1.74) 3.84 (3.12–4.55) 1.73 (1.56–1.90) 14.70 (13.48–15.46) 1.53 (1.40–1.65) 6.72 (5.85–7.59) 1.34 (1.21–1.47)  Primary health care 1.12 (0.87–1.36) 1.76 (1.62–1.90) 1.91 (1.60–2.21) 1.80 (1.68–1.91) 1.73 (1.37–2.09) 1.65 (1.54–1.75) 1.32 (1.04–1.56) 1.45 (1.35–1.56) Professionals met on visits  Psychologist, therapist, social worker 0.46 (0.21–0.70) 0.35 (0.29–0.41) 0.76 (0.53–0.95) 0.35 (0.29–0.41) 3.16 (2.68–3.63) 0.32 (0.27–0.38) 1.17 (0.86–1.48) 0.28 (0.22–0.33)  Physician 2.31 (1.73–2.89) 2.34 (2.28–2.52) 3.88 (3.23–4.53) 2.44 (2.32–2.56) 8.09 (7.50–8.68) 2.20 (2.09–2.31) 4.28 (3.71–4.84) 1.91 (1.80–2.02)  Other 3.21 (1.29–5.12) 1.73 (1.46–1.99) 4.41 (2.51–6.31) 1.88 (1.62–2.15) 9.31 (6.76–11.85) 1.72 (1.47–1.96) 5.60 (3.67–7.53) 1.3727 (1.2128–1.5327)

HIV = human immunodeficiency virus; CI = confidence interval.

Comparisons in all men with new HIV diagnosis compared with matched HIV-negative controls 2 years (366–730 days) and 1 year (1–365 days) before diagnosis and 1 year (1–365 days) and 2 years (366–730 days) after the first registered HIV diagnosis, during the period 2011–2018. Significant differences were defined as no overlap in the 99% CIs between cases and controls at each time point and are shown in bold.

No statistical tests were made. Nonoverlapping 99% CIs were considered significant.


Health Care Visits in Women and Men Before the First Registered Diagnosis of HIV

Women with HIV had fewer visits to specialist clinics, to primary care, and to physicians in general 2 years before their HIV diagnosis than their matched controls. Women with HIV also had fewer visits to primary care 1 year before HIV diagnosis than their controls.

Men with HIV visited primary care less 2 years before HIV diagnosis. Men visited all clinics except primary care significantly more than their controls 1 year before HIV diagnosis. Similarly, there was no difference in the contacts with health care professionals 2 years before diagnosis, but men with HIV had significantly more visits to all health care professionals than their controls 1 year before diagnoses.

Health Care Visits in Women and Men After the First Registered Diagnosis of HIV

Women with HIV visited outdoor specialist clinics, gynecologists, and other specialist clinics more than their controls 1 year after HIV diagnosis and visited the same clinics more than their controls 2 years after HIV diagnosis. Women visited primary care significantly less than their controls 2 years after HIV diagnosis. Women with HIV visited all health care professionals significantly more than their controls 1 and 2 years after diagnosis.

Men with HIV undertook significantly more visits to all clinics except primary care 1 and 2 years after diagnosis. They also visited all health care professionals more than their controls after diagnosis.

Psychiatric comorbidities and collected drug prescriptions are shown in Table 4 for women and Table 5 for men, and the results are summarized hereinafter.

TABLE 4 - ORs for Relevant Psychiatric Comorbidities and Drug Prescriptions of Relevant Pharmacotherapies in Women With HIV Diagnosis Compared With Matched HIV-Negative Controls 2 Years (366–730 Days) and 1 Year (1–365 Days) Before Diagnosis and 1 Year (1–365 Days) and 2 years (366–730 Days) After HIV Diagnosis (2011–2018) 2 y Before Diagnosis (n = 450), OR (99% CI) 1 y Before Diagnosis (n = 450), OR (99% CI) 1 y After Diagnosis (n = 450), OR (99% CI) 2 y After Diagnosis (n = 450), OR (99% CI) Diagnoses  Suicide attempt (X60–84) 9.75 (1.19–80.18)*** 2.43 (0.32–18.69) 2.78 (0.35–21.99) 1.62 (0.10–26.19)  Substance use disorder (F11–F19) 5.29 (1.57–17.80)*** 6.05 (1.89–19.40)*** 5.24 (1.69–16.32)*** 3.49 (0.91–13.44)  Alcohol use disorder (F10) 1.87 (0.53–6.53) 3.46 (1.01–11.82)** 2.43 (0.75–7.97) 1.95 (0.55–6.93)  Depression (F32, F33) 0.84 (0.43–1.63) 0.86 (0.46–1.59) 0.94 (0.50–1.77) 1.09 (0.55–2.17)  Fatigue syndrome (F43.8) 0.68 (0.14–3.16) 0.38 (0.06–2.43) 0.41 (0.06–2.64) 0.35 (0.06–2.24)  Stress (F43) 0.52 (0.19–1.43) 0.63 (0.27–1.46) 1.20 (0.624–2.319) 1.52 (0.73–3.13)  Anxiety (F40, F41) 0.59 (0.28–1.26) 0.60 (0.30–1.20) 0.85 (0.47–1.55) 1.14 (0.61–2.12)  Bipolar disorder (F30, F31) 0.40 (0.029–5.60) 0.77 (0.12–5.16) 2.44 (0.70–8.90) 1.95 (0.47–8.03)  Psychosis (F20, F23, F25, F28, F29) 0.39 (0.03–5.36) 0.81 (0.12–5.39) 0.35 (0.03–4.75) 1.12 (0.23–5.41) Pharmaceutical drug type  Stimulants (N06BA) 1.33 (0.43–4.10) 0.95 (0.28–3.20) 1.42 (0.50–4.06) 1.43 (0.46–4.43)  Neuroleptics (N05A) 1.17 (0.49–2.80) 1.66 (0.80–3.44) 2.37 (1.23–4.56)*** 1.68 (0.79–3.57)  Benzodiazepines (N05BA) 0.80 (0.35–1.81) 0.98 (0.46–2.09) 1.27 (0.60–2.70) 0.84 (0.34–2.09)  Tranquilizers (R06AD02, N05BB) 0.77 (0.41–1.44) 1.06 (0.63–1.81) 1.12 (0.66–1.91) 0.95 (0.51–1.79)  Hypnotics (N05CF) 0.77 (0.39–1.48) 0.72 (0.38–1.37) 1.21 (0.71–2.07) 1.22 (0.69–2.17)  Propiomazine (N05CM06) 0.58 (0.19–1.71) 0.99 (0.43–2.25) 1.33 (0.65–2.74) 1.11 (0.44–2.76)  Antidepressant drugs (N06A) 0.46 (0.26–0.83)*** 0.66 (0.08–1.06) 0.87 (0.57–1.33) 0.76 (0.46–1.24)

OR = odds ratio; HIV = human immunodeficiency virus; CI = confidence interval.

Conditional logistic regression. Significant differences were defined as p < .01 and are shown in bold.

** p < .01.

*** p < .001.


TABLE 5 - ORs for Relevant Psychiatric Comorbidities and Drug Prescriptions of Relevant Pharmacotherapies in Men With HIV Diagnosis Compared With Matched HIV-Negative Controls 2 Years (366–730 Days) and 1 Year (1–365 Days) Before Diagnosis and 1 Year (1–365 Days) and 2 Years (366–730 Days) After HIV Diagnosis (2011–2018) 2 y Before Diagnosis (n = 930), OR (99% CI) 1 y Before Diagnosis (n = 930), OR (99% CI) 1 y After Diagnosis (n = 930), OR (99% CI) 2 y After Diagnosis (n = 930), OR (99% CI) Diagnoses  Suicide attempt (X60–84) — a 3.23 (0.17–63.32) 0.97 (0.07–14.45) 2.42 (0.14–43.20)  Substance use disorder (F11–F19) 2.54 (1.13–5.69)** 4.36 (2.00–9.51)*** 5.16 (2.65–10.08)*** 3.86 (1.65–9.01)***  Stress (F43) 1.15 (0.50–2.62)

留言 (0)

沒有登入
gif