metricas
covid
Buscar en
Spanish Journal of Psychiatry and Mental Health
Toda la web
Inicio Spanish Journal of Psychiatry and Mental Health Epidemiological characteristics and hospitalization trajectories prior to suicid...
Información de la revista
Vol. 16. Núm. 2.
Páginas 76-84 (abril - junio 2023)
Compartir
Compartir
Descargar PDF
Más opciones de artículo
Visitas
735
Vol. 16. Núm. 2.
Páginas 76-84 (abril - junio 2023)
Original article
Acceso a texto completo
Epidemiological characteristics and hospitalization trajectories prior to suicide in Galicia between 2013 and 2016
Visitas
735
Vanessa Blancoa,
Autor para correspondencia
vanessa.blanco@usc.es

Corresponding author.
, María Tajes Alonsob, Luisa F. Peleteiro Pensadob, Gael Naveira Barbeitob, Daniel Núñez Ariasb, Ángela J. Torresc, Manuel Arrojoc, Mario Páramoc, Patricia Oterod, Fernando L. Vázqueze
a Department of Evolutionary and Educational Psychology, University of Santiago de Compostela, Santiago de Compostela, Spain
b Galician Healthcare Service (SERGAS), Santiago de Compostela, Spain
c Department of Psychiatry, Radiology, Public Health, Nursing and Medicine, University of Santiago de Compostela, Santiago de Compostela, Spain
d Department of Psychology, University of A Coruña, A Coruña, Spain
e Department of Clinical Psychology and Psychobiology, University of Santiago de Compostela, Santiago de Compostela, Spain
Este artículo ha recibido
Información del artículo
Resumen
Texto completo
Bibliografía
Descargar PDF
Estadísticas
Figuras (1)
Tablas (5)
Table 1. Sociodemographic, clinical, and forensic profile of people who died by suicide in Galicia between 2013 and 2016 (n=1354).
Table 2. Suicide mortality rates (SMR) by year, sex, age group and area.
Table 3. Associations of sociodemographic and medical variables for each group of trajectories according to mean number of hospitalizations (n=1354).
Table 4. Association between the different hospitalization trajectories prior to suicide and the demographic and clinical adjustment variables (multinomial model).
Table 5. Association between the different trajectories and the covariates (multinomial model).
Mostrar másMostrar menos
Abstract
Introduction

Addressing suicide requires an understanding of regional patterns of epidemiology, with health variables being central. However, the clinical profile of people who commit suicide has received little attention. The objectives of this study were to analyze the sociodemographic, clinical, and forensic characteristics of persons who committed suicide in Galicia between 2013 and 2016, analyze suicide mortality rates, and identify trajectories of hospitalizations and associated variables.

Material and methods

A population study was carried out on the 1354 people who died by suicide in Galicia.

Results

The most common profile was a retired man, 57.9 years old (SD=18.5), from an urban and inner area. 43.6% had been previously hospitalized, 41.6% had been diagnosed with physical disorders, and 26.8% with mental disorders. 48.2% had been prescribed psychiatric medications and 29.6% had received outpatient psychiatric care. The highest prevalence of death by suicide (27.5%) was in 2014, with the predominant method being hanging (59.1%). The average raw rate was 12.3/100,000. Three trajectories of hospitalizations emerged: 94.83% had experienced few hospitalizations; 2.95% an increasing pattern; and 2.22% a decreasing pattern. These trajectories were associated with number of psychiatric appointments, prescription of psychiatric medications, and diagnoses of physical and mental disorders.

Conclusions

These findings are crucial for detection and prevention.

Keywords:
Clinical variables
Epidemiology
Profile
Psychiatric appointments
Psychiatric medications
Physical disorders
Mental disorders
Texto completo
Introduction

Completed suicides represent a major public health problem. Globally, an estimated 793,000 people died by suicide in 2016, which is equivalent the normalized suicide rate of 10.5 per 100,000 population, 12.9 per 100,000 in Europe.1 Beyond these figures, each case represents a personal tragedy and has a devastating impact on the people close to the deceased. The public health system suicide prevention model highlights the importance of vigilance as part of a culturally specific and data-supported comprehensive national strategy; despite this, suicide has remained a low public health priority.2 This is especially relevant in Galicia (an autonomous community in northwest Spain) because although trends in regional suicide rates have decreased slightly between 1988 and 2012,3 until then they were still among the highest in the nation.4,5 Updated statistics on suicide in this region would facilitate adoption of tailored prevention strategies aimed at decreasing the rate of suicide and the suffering associated with it.

Most national epidemiological studies of suicide (e.g., Refs. 3–5) analyze the sociodemographic profile of patients but do not emphasize health-related variables. Moreover, at the international level, studies examining clinical variables are also few and far between, and have focused on context-specific aspects of the clinical variables. Though this type of analysis is useful, the generalizability is limited. Studies of persons who have completed suicide have found the percentage diagnosed with physical disorders ranges between 27%6 and 55%.7 Likewise, mental disorders have been related to suicidal behavior, particularly substance and alcohol use,8 and mood disorders.9,10 One study found that 33% of people who died by suicide had suffered from depression, while 47% had depressive symptoms.7 Furthermore, 42.6% were taking psychiatric medications.11 There is limited evidence on outpatient psychiatric care, but one study12 found that around 60% of completed suicides occurred in patients who were receiving it.

Regarding hospitalization rates, hospital admissions due to physical illnesses were associated with an increased risk of suicide, that was higher in patients with three or more diagnoses and highest for those who had been hospitalized in the previous two years.13 Two studies provided data on the year prior to the suicide, finding hospitalization rates of 5.7%14 and 27.7%.15 Those who had completed suicide were more likely to have had more hospital admissions for psychiatric services.16 One study11 reported that 14.2% had experienced a psychiatric hospitalization in the previous year, while another study17 found that 31.2% had been hospitalized for mental health services during the seven preceding years. Likewise, it has been found that sociodemographic variables such as age (e.g., Ref 13), and clinical variables such as prescriptions for psychiatric medications (e.g. Ref. 18) interact with hospitalizations in predicting the risk of suicide completion. Though this limited previous evidence suggests that a subgroup of people who have died by suicide have been hospitalized previously, and these hospitalizations may be related to increased risk of suicide, no previous research has examined hospitalization rates and trajectories prior to suicide or the relationship between sociodemographic and clinical variables and these trajectories.

Knowing the clinical profile of patients who die by suicide, including the trajectories of hospitalizations prior to suicide, is crucial for the development of indicators that can help clinicians identify persons at risk; moreover, contact with health services during hospitalizations prior to suicide would provide the clinicians with the opportunity to adopt the necessary preventive measures.

The objectives of this study were (a) to analyze the sociodemographic, clinical, and forensic characteristics of the people who died by suicide in Galicia between 2013 and 2016; (b) to analyze suicide mortality trends for that period in terms of sex, age, and area; and (c) identify the hospitalization trajectories prior to suicide and their associations with sociodemographic and clinical variables.

Materials and methodsDesign and participants

This population study included all deaths in Galicia between January 2013 and December 2016 whose medical/legal etiology was deemed to be death by suicide after a forensic autopsy. Galicia is an autonomous community, consisting of four provinces further divided into 313 municipalities. Its total area is 29,574km2, and it is home to 2,699,499 residents.19

The study used secondary data obtained from the Galician Mortality Registry (RMG), produced by the Epidemiology Service of the Galician Healthcare Service (SERGAS) related to deaths occurred between 2013 and 2016 when the cause of death was suicide (codes X60–X84 and Y87.0 of the International Classification of Diseases, 10th revision20; data on sex, age, year and manner of death). Likewise, the researchers accessed the SERGAS Minimum Basic Set of Hospital Discharges (CMBD-AH) database produced by the Care Evaluation and Guarantees Service to obtain information regarding hospital admissions of the people identified in the RMG registry (data on whether they had been hospitalized, number of hospitalizations, discharging service, and hospital diagnoses between 2010 and 2016); the Health Card database from the Health Card and Personal Accreditation Service (data on the employment situation); the Nasi Psychiatry database of the Care Evaluation and Guarantees Service (data on attendance at outpatient psychiatric care in the public sector and number of sessions for 2013–2016) and the Psychopharmaceuticals database of the Secretariat of the General Pharmacy Subdivision (data on prescriptions of psychiatric medications and type between 2014 and 2016).

Patient records/information were anonymized. The study was approved by the Santiago-Lugo Research Ethics Committee, SERGAS (Code: 2018/135).

Statistical analysis

An exploratory analysis was performed with the raw data. Descriptive statistics were calculated. Regarding mental and physical disorders, only those with a prevalence of 5% or higher were included to simplify the presentation of the data. Crude suicide mortality rates between 2013 and 2016 were calculated per 100,000 inhabitants using the demographic data of Galicia obtained from the Galician Institute of Statistics21 and further categorized by sex and age group, for each year. These rates represent the real rate.22 Likewise, age-adjusted rates were calculated using the WHO World Standard Population23 to make comparisons with other studies.22 To calculate rates by age groups, age was categorized into: <15, 15–24, 25–34, 35–44, 45–54, 55–64, 65–74, 75–84, ≥85.

Group-based trajectory modeling was used to estimate pre-suicide hospitalization trajectories from T-3 (3 years before the suicide) to T-0 (the same year as the suicide). This procedure is based on a mixed model that allows the identification of subgroups of individuals who follow different trajectories during the period and estimates a regression model for each group.24,25 The following steps were used: (a) determine the type of random structure and the need for flexibility in the predictor; (b) determine the optimal number of classes, with each class including at least 1% of the population (based on the evaluation of the adjustment criteria by Klijn et al.26); and (c) re-estimate the optimal structure of the model with the optimal number of classes.

Finally, the associations among the sociodemographic variables, mental and physical disorder diagnoses, medication use, and number of psychiatric appointments were estimated using Chi-square, Student's t-test, and multinomial logistic regression for each of the identified groups, excluding patients with missing values. Chi-square likelihood ratio tests were used to assess whether the variables were associated with the type of trajectory in the full model, and Nagelkerke's R2 to assess the strength of these associations. Given the high collinearity among the variables, each factor was excluded from the full model in sequence, and the differences in R2 were then calculated for each to explore the contribution of each factor to the full model. The model for each of the variables is thus: Ysex+age+no. of psychiatric appointments+psychiatric medications+X (where X is each of the variables).

All statistical analyses were carried using IBM's SPSS Statistics software for Windows, version 24.0 and the freeware program R. The packages lcmm27 and LCTMtools28 were used to estimate the trajectories, and nnet29 to adjust the multinomial models.

ResultsCharacteristics of people who died by suicide during the period

Between 2013 and 2016, 1354 completed suicides occurred in Galicia, though the most occurred in 2014 (27.8% of cases). As shown in Table 1, 72.8% were men, with a mean age of 57.9 years old (SD=18.5), ranging from 12 to 95. The majority lived in urban areas (62.8%) in the interior of the country (58.3%) and were retired (49.2%). The method of death by suicide was a non-toxic violent agent in 92.0%; hanging, strangulation, or suffocation for 59.1%.

Table 1.

Sociodemographic, clinical, and forensic profile of people who died by suicide in Galicia between 2013 and 2016 (n=1354).

Variable  n 
Sex
Male  986  72.8 
Female  368  27.2 
Age
M  57.9   
SD  18.5   
Rank  12–95   
Rural vs. urban area
Rural  504  37.2 
Urban  850  62.8 
Interior vs. coastal area
Interior  789  58.3 
Coast  565  41.7 
Employment status
Active worker  287  21.2 
Retiree  666  49.2 
Unemployed/other  267  19.7 
Unknown/not reported  134  9.9 
Hospitalized between 2010 and 2016
No  764  56.4 
Yes  590  43.6 
Number of times hospitalized between 2013 and 2016
2.2   
SD  2.0   
Rank  1–18   
Discharging service*
Pediatrics  0.1 
Ophthalmology  0.4 
Digestive and General and Digestive Surgery  133  9.8 
Internal Medicine  155  11.4 
Pulmonology and Thoracic Surgery  38  2.8 
Traumatology  67  4.9 
Mental Health  180  13.3 
Neurology and Neurosurgery  47  3.5 
Nephrology  0.6 
Intensive Care  43  3.2 
Urology  40  3.0 
Oncology  13  1.0 
Rehabilitation  0.1 
Cardiology, Cardiac Surgery, Angiology, and Vascular Surgery  57  4.2 
Geriatrics  0.2 
Hematology  0.4 
Discharging service*
Gynecology and Obstetrics  15  1.1 
Rheumatology  0.3 
Endocrinology  0.3 
Dermatology  0.2 
Anesthesiology and Resuscitation  11  0.8 
Plastic and Reconstructive Surgery  0.4 
Dentistry and Maxillofacial Surgery  0.2 
Otorhinolaryngology  16  1.2 
Mental disorders363  26.8 
Mood disorders  159  11.7 
Depression  145  10.7 
Substance-related disorders and addictive disorders  179  13.2 
Personality disorders  75  5.5 
Physical disorders563  41.6 
Cardiovascular diseases  283  20.9 
Hematological diseases  117  8.6 
Respiratory diseases  204  15.1 
Endocrine diseases  234  17.3 
Musculoskeletal diseases  184  13.6 
Neurological diseases  160  11.8 
Digestive system diseases  221  16.3 
Diseases of the genitourinary system  148  10.9 
Infectious or parasitic diseases  69  5.1 
Wounds, injuries, and adverse effects of medication or medical treatments  234  17.3 
Intentionally self-inflicted injuries  145  10.7 
Factors influencing health status  280  20.7 
Received prescription for psychoactive drug653  48.2 
Stimulants  0.5 
Antidepressants  465  34.3 
Mood stabilizers  13  1.0 
Anxiolytics and sedatives/hypnotics  569  42.0 
Antipsychotics  267  19.7 
Has attended a psychiatric service between 2013 and 2016
No  953  70.4 
Yes  401  29.6 
Number of psychiatric appointments between 2013 and 2016
M  6.4   
SD  7.4   
Rank  1–50   
Have had contact with any of the health services analyzed  921  68.0 
Year of death
2013  330  24.4 
2014  376  27.8 
2015  309  22.8 
2016  338  25.0 
Method of suicide
Toxic agent  108  8.0 
Other drugs and alcohol  50  3.7 
Antiepileptic, sedative, hypnotic, or antiparkinsonian drug  26  1.9 
Gas or vapor  15  1.1 
Other chemicals and noxious substances  11  0.8 
Narcotics and psychodysleptics  0.5 
Violent non-toxic agent  1246  92.0 
Hanging, strangulation, or suffocation  800  59.1 
Jumping from height  210  15.5 
Drowning and submersion  86  6.3 
Firearm  65  4.8 
Sharp object  26  1.9 
Moving object/vehicular collision  16  1.2 
Smoke, fire, and flame  0.3 
Other means  39  2.9 

Note: *Response categories are not mutually exclusive.

Regarding the clinical characteristics, 43.6% had been hospitalized between 2010 and 2016, for an average of 2.2 times (SD=2.0). 68.0% had some hospitalization or contact with psychopharmaceutical or psychiatric services. Most of the previous discharges had been from mental health services (13.3%). Of the diagnoses, 26.8% had been mental health disorders (the most frequent being substance use disorders, at 13.2%) and 41.6% physical disorders (the most frequent being cardiovascular diseases, at 20.9%). 48.2% had been prescribed some psychiatric medication—anxiolytics and sedatives/hypnotics in 42.0%; 29.6% had received psychiatric services between 2013 and 2016, with an average of 6.4 visits (SD=7.4).

Suicide mortality rate per 100,000 population

Table 2 shows suicide mortality rates (SMR) per 100,000 population. The average raw SMR during the period was 12.3. The year with the highest SMR was 2014 (13.7).

Table 2.

Suicide mortality rates (SMR) by year, sex, age group and area.

  2013  2014  2015  2016  Average 
Gross rate  12.0  13.7  11.3  12.4  12.3 
Sex
Male  18.2  21.6  15.7  16.6  18.0 
Female  6.6  7.0  6.0  7.3  6.7 
Age group (years)
<15  0.3  0.0  0.0  0.3  0.2 
15 and 24  4.7  7.0  4.5  6.8  5.7 
25 and 34  7.2  7.3  5.9  6.9  6.8 
35 and 44  8.1  14.1  10.8  13.7  11.7 
45 to 54  18.5  18.2  13.9  13.7  16.1 
55 to 64  14.6  14.0  14.7  12.5  13.9 
65 to 74  18.3  19.7  16.8  17.0  17.9 
75 to 84  26.0  26.2  24.4  25.3  25.5 
85 or older  13.0  25.0  12.0  24.0  18.5 
Rural/urban area
Rural  14.7  20.0  14.3  17.9  16.7 
Urban  10.9  11.3  10.2  10.4  10.7 
Adjusted rate (WHO)  8.1  9.3  7.5  8.4  8.3 

Note: The raw rate refers to the number of suicides per 100,000 population divided by the reference population; the Adjusted Rate (WHO) refers to the same data weighted according to the WHO Standard World Population structure (Ahmad et al., 2001).

SMR by sex and age. The SMR was higher in men, with 18.0 suicides per 100,000 men (ranging from 15.7 in 2015 to 21.6 in 2014), compared to 6.7 per 100,000 women (ranging between 6.0 in 2015 and 7.3 in 2016). Regarding age, the highest SMR occurred among persons ≥65 years. Specifically, the age group with the highest average SMR was 75–84 years (25.5). Finally, the highest SMR corresponded to rural areas (16.7).

Adjusted SMR. Using the world population as a reference,23 the average rate was 8.3 per 100,000, ranging between 7.5 for 2015 and 9.3 for 2014.

Hospitalization trajectories and relationships between suicide and sociodemographic and clinical variables

Trajectories. Three different groups of hospitalization trajectories were identified as the best-fit model (see Fig. 1). These three groups were labeled: Group 1: Increasing, Group 2: Low Constant, and Group 3: Decreasing.

Fig. 1.

Hospitalization trajectories from three years prior to suicide to one year after suicide, mean number of hospitalizations per year.

(0.31MB).

The Increasing group included 2.95% of the participants (n=40). It had a low number of hospitalizations 3 years prior to the suicide, a slight increase up to 2 years before, a more abrupt increase between 2 and 1 year before, and a continual gradual increase between the year before the suicide and the year of suicide. A large proportion of the participants (94.83%; n=1284) belonged to the Low constant group; they consistently had few or no hospitalizations in the years prior to suicide. The Decreasing group included 2.22% of the participants (n=30). It began with a high number of hospitalizations between 3 and 2 years before suicide, experienced a slight decrease between 2 and 1 year before, and a more marked decrease between the year before and the year of the suicide.

Association of sociodemographic and clinical variables.Table 3 presents the contingency tables for each of the sociodemographic and clinical variables with each trajectory.

Table 3.

Associations of sociodemographic and medical variables for each group of trajectories according to mean number of hospitalizations (n=1354).

Variable  Groupp 
  Increasing (n=40)  Low constant (n=1284)  Decreasing (n=30)   
Sex0.704 
Male  27 (67.5%)  938 (73.1%)  21 (70.0%)   
Female  13 (32.5%)  346 (26.9%)  9 (30.0%)   
Age0.060 
Mean (SD62.52(17.08)  57.59 (18.55)  63.53 (17.25)   
Number of psychiatry appointments0.002 
Mean (SD4.03 (7.61)  1.79 (4.73)  3.93 (8.64)   
Prescription of psychiatric medications<0.001 
No  9 (22.5%)  676 (52.6%)  16 (53.3%)   
Yes  31 (77.5%)  608 (47.4%)  14 (46.7%)   
Mood disorders<0.001 
No  24 (60.0%)  1152 (89.7%)  19 (63.3%)   
Yes  16 (40.0%)  132 (10.3%)  11 (36.7%)   
Substance-related and addictive disorders<0.001 
No  23 (57.5%)  1139(88.7%)  13 (43.3%)   
Yes  17 (42.5%)  145 (11.3%)  17 (56.7%)   
Personality disorders<0.001 
No  31 (77.5%)  1224 (95.3%)  24 (80.0%)   
Yes  9 (22.5%)  60 (4.7%)  6 (20.0%)   
Cardiovascular diseases<0.001 
No  16 (40.0%)  1046 (81.5%)  9 (30.0%)   
Yes  24 (60.0%)  238 (18.5%)  21 (70.0%)   
Hematological diseases<0.001 
No  25 (62.5%)  1197 (93.2%)  15 (50.0%)   
Yes  15 (37.5%)  87 (6.8%)  15 (50.0%)   
Respiratory diseases<0.001 
No  14 (35.0%)  1126 (87.7%)  10 (33.3%)   
Yes  26 (65.0%)  158 (12.3%)  20 (66.7%)   
Endocrine diseases<0.001 
No  14 (35.0%)  1098 (85.5%)  8 (26.7%)   
Yes  26 (65.0%)  186 (14.5%)  22 (73.3%)   
Musculoskeletal diseases<0.001 
No  23 (57.5%)  1132 (88.2%)  15 (50.0%)   
Yes  17 (42.5%)  152 (11.8%)  15 (50.0%)   
Neurological diseases<0.001 
No  23 (57.5%)  1152 (89.7%)  19 (63.3%)   
Yes  17 (42.5%)  132 (10.3%)  11 (36.7%)   
Digestive system diseases<0.001 
No  23 (57.5%)  1101 (85.7%)  9 (30.0%)   
Yes  17 (42.5%)  183 (14.3%)  21 (70.0%)   
Diseases of the genito-urinary system<0.001 
No  25 (62.5%)  1163 (90.6%)  18 (60.0%)   
Yes  15 (37.5%)  121 (9.4%)  12 (40.0%)   
Infectious diseases<0.001 
No  30 (75.0%)  1236 (96.3%)  19 (63.3%)   
Yes  10 (25.0%)  48 (3.7%)  11 (36.7%)   
Wounds, injuries, and adverse effects of medications and treatments<0.001 
No  14 (35.0%)  1100 (85.7%)  6 (20.0%)   
Yes  26 (65.0%)  184 (14.3%)  24 (80.0%)   
Intentionally self-inflicted injuries<0.001 
No  23 (57.5%)  1165 (90.7%)  21 (70.0%)   
Yes  17 (42.5%)  119 (9.3%)  9 (30.0%)   
Factors influencing health status<0.001 
No  10 (25.0%)  1057 (82.3%)  7 (23.3%)   
Yes  30 (75.0%)  227 (17.7%)  23 (76.7%)   

Sex and age were not significantly associated with the groups of hospitalization trajectories.

Table 4 contains the results for the association between the different trajectories and the sociodemographic and clinical variables for adjustment of the model (sex, age, number of psychiatric appointments, psychiatric medications).

Table 4.

Association between the different hospitalization trajectories prior to suicide and the demographic and clinical adjustment variables (multinomial model).

Variables  LR.Chisq  Pr.Chisq.  R2 
Sex  0.704  0.7031  0.1369835 
Age  5.750  0.0564  1.1160176 
Number of psychiatry appointments  8.849  0.0120  1.7154475 
Prescription of psychiatric medications  14.833  0.0006  2.8692704 

Note: LR.Chisq: Chi-Square goodness of fit; Pr.Chisq: Chi-Square probability; R2 (expressed as a percentage).

The categories of disorders analyzed were added to this model, one by one. Table 5 presents the results of the multinomial models used to evaluate associations between each of these variables and the hospitalization trajectories.

Table 5.

Association between the different trajectories and the covariates (multinomial model).

Variables  LR.Chisq  Pr.Chisq.  chisq  p  R2  Diff 
Mood disorders  25.985  0.0000  35.888  0.0000  10.335638  3.447132 
Substance-related and addictive disorders  65.288  0.0000  56.543  0.0000  17.574426  6.803369 
Personality disorders  21.602  0.0000  22.331  0.0000  9.515280  5.207619 
Cardiovascular diseases  56.023  0.0000  66.588  0.0000  15.886927  3.249016 
Hematological diseases  60.132  0.0000  65.709  0.0000  16.636763  4.161548 
Respiratory diseases  88.440  0.0000  100.056  0.0000  21.741367  2.982268 
Endocrine diseases  83.427  0.0000  97.636  0.0000  20.845092  2.523528 
Musculoskeletal diseases  38.304  0.0000  46.184  0.0000  12.627252  3.796002 
Neurological diseases  30.271  0.0000  39.208  0.0000  11.135235  3.618757 
Digestive system diseases  55.258  0.0000  62.140  0.0000  15.747146  3.934130 
Genitorurinary diseases  32.241  0.0000  39.309  0.0000  11.501935  3.966368 
Infectious diseases  50.457  0.0000  51.102  0.0000  14.867620  5.113494 
Wounds, injuries, and adverse effects of medications and treatments  103.873  0.0000  109.534  0.0000  24.479649  4.014367 
Intentionally self-inflicted injuries  35.681  0.0000  37.847  0.0000  12.141037  4.881787 
Factors influencing health status  91.912  0.0000  104.673  0.0000  22.360115  2.768317 

Note: LR.Chisq=Chi-Square goodness of fit; Pr.Chisq=Chi-Square probability; chisq=Chi-square; R2=Nagelkerke's R2 for the full model (expressed as a percentage); Diff=difference between the Nagelkerke's R2 for the full model and Nagelkerke's R2 for the model that includes only the covariate.

Differences in R2 indicated that endocrine diseases (diff in R2=2.523528) played the most important role in the full model. Furthermore, in comparison to the Low Constant group a higher average number of psychiatric appointments was found in the Increasing group. In addition, those who used psychiatric medications were overrepresented, as were persons who suffered from the following disorders: mood, substance-related, personality; cardiovascular disease; diseases of the blood, respiratory system, endocrine system; musculoskeletal disorders; neurological disorders; disorders of the digestive or the genitourinary system; infectious diseases; wounds, injuries, and adverse effects from medications and medical treatments; intentionally self-inflicted injuries and factors influencing health status. The pattern is similar in the Decreasing group in comparison to the Low Constant group, with the exception of no difference in prescriptions for psychiatric medications. The Low constant group had fewer psychiatric appointments and a lower incidence of mental and physical pathologies than the other two groups.

Discussion

The aims of the current study were to analyze the profile of people who died by suicide in Galicia between 2013 and 2016; to analyze suicide mortality trends; and to identify the hospitalization trajectories prior to suicide and their associations with sociodemographic and clinical variables.

The sociodemographic profile of people who died by suicide was a retired man, 58 years old, from an urban area in the interior of the country. Previous studies have also found that the largest group of people who had died by suicide in Galicia were men (e.g., Refs. 3–5), between the ages of 45 and 65 (e.g., Ref. 3), with limited economic resources (e.g., Refs. 3, 4). Likewise, other studies outside the region (e.g., Refs. 30, 31) found a higher number of cases between the retirees. Though the finding of an urban profile may seem to contrast with previous studies that found higher suicide rates in rural areas (e.g., Ref. 4), this divergence is due to the fact that we analyzed profiles of people who died by suicide, not rates as a function of the type of living area. Urban areas are much more populated in our region; in fact, our results regarding rates confirmed higher SMR in rural areas.

Regarding clinical profiles, 43.6% had been hospitalized between 2010 and 2016. Hospitalizations were lower than those found in a study that evaluated the two years prior to suicide,13 but much higher than those of two investigations focused on the previous year14,15; the diversity of time frames and cultural backgrounds, together with the diverse ages and physical conditions of the samples used, makes comparing the data difficult. Specifically, 13.3% were hospitalized for mental health problems, which was consistent with Paraschakis et al.11

Regarding diagnoses during hospitalizations, 41.6% had at least one physical illness, predominantly cardiovascular disease, and 26.8% a mental disorder, predominantly related to substance use or mood disorders. The prevalence of diagnoses for cardiovascular diseases, substance and alcohol use, and mood disorders was consistent with previous studies (e.g., Refs. 6, 8, 10), although our figures for mental disorders were lower, likely because outpatient diagnoses were excluded. Almost half had received a prescription for psychiatric medications, especially anxiolytics and sedatives/hypnotics (42.0%), and almost one-third had attended psychiatric care through the public health service. Our participants were twice as likely to have been prescribed anxiolytics than the figures reported in the general adult population (18.4%32). However, the prevalence of psychiatric care visits was half of that found by Lehman et al.12 in the United States. The differences between countries in the use and availability of mental health services and our study's exclusion of private providers might explain these discrepancies. Altogether, more than two-thirds had contact with one of these services in the years before the death, which represents a total number of contacts with health services three times higher than those that Basham et al.,14 found for the year prior to suicide; and similar to that found by Cheung et al.7 for people who died by suicide during late life, where 6.9 to 51% had consulted their general practitioners in the month before they died by suicide and 9.6–15% had contacted psychiatric services within 1 month. Though these data are very high, it is noteworthy that our study did not include consults to general practitioners, the health professionals closest to the patients in our context.

In relation to the forensic profile, the year with the most deaths was 2014, a finding consistent with the Spanish data.33 The predominant method was a non-toxic violent agent; specifically, 59.1% completed suicide by hanging, strangulation, or suffocation, in line with previous international studies (e.g., Refs. 6, 11, 34).

In regard to suicide mortality trends, the average gross SMR in Galicia for the entire period was 12.3, much higher than the Spanish rate of 7.6 per 100,000 population.33 The highest SMR corresponded to 2014, to men between the ages of 75 and 84 who are residents of rural areas. The highest rates in terms of year, sex, and age group are consistent with a previous study in Galicia,3 and the finding that rates were highest in rural areas reiterates previous national (e.g., Ref. 4) and international studies (e.g., Refs. 35, 36). Regarding the adjusted rates, the average rate standardized based on the world population was 8.3, a figure lower than global and European rates.1

With respect to the hospitalization trajectories prior to suicide, we identified three groups of trajectories according to the mean number of hospitalizations: Increasing (Group 1), Low constant (Group 2), and Decreasing (Group 3). The Increasing group (almost 3% of the participants) had an increase in hospitalizations starting 3 years prior to the suicide, that became more pronounced between 2 and 1 years prior. The Decreasing group (over 2%) reported a high number of hospitalizations three years before the suicide and followed a trend of reduction up until the year of the suicide. The majority (almost 95%) belonged to the Low Constant group, characterized by a low number of hospitalizations in previous years. This three-group finding confirms the well-established belief that suicide victims may exhibit different patterns of contact with health care services and underlying illnesses.37 Though the figures of people who died by suicide with atypical trajectories of hospitalizations are low, the finding that 5% of the people who died by suicide exhibited a characteristic pattern consisting of an increasing or decreasing number of hospitalizations during the three years prior to suicide represents an opportunity to intensify the measures for the detection and prevention of suicide among patients with these clinical trajectories.

Finally, the factors associated with these groups were: number of psychiatric appointments; prescription of psychiatric medications; and a series of mental disorders (mood, substance-related and addictive, personality) and physical diseases (cardiovascular; hematological; respiratory; endocrine; musculoskeletal; neurological; digestive; genitourinary; infectious; wounds, injuries, and adverse effects from medications and medical treatments; intentionally self-inflicted injuries; and factors influencing health status). Endocrine diseases had the greatest contribution to the full model. The Increasing group had more contact with psychiatric services, had been prescribed more psychiatric medications, had more mental and physical disorders, a history of intentional self-harm, and had experienced various factors with negative repercussions on their state of health. The profile for the Decreasing group was similar, but with lower rates for the prescription of psychiatric medications, coupled with higher rates of substance use and physical illness and a very high percentage of wounds, injuries, and adverse effects from medication and medical treatments. Mental and physical pathologies and psychiatric consultations were lower among the Low Constant. A tentative hypothesis to these findings is that the increasing group could be composed of people who suffer from acute psychiatric conditions and do not achieve remission, while the decreasing group could consist of people who have many physical problems and end up dying by suicide because they find no solution. The finding that both groups were associated with high rates of physical comorbidity is consistent with previous studies8,10,38 that found that physical conditions, especially when they are chronic and linked to pain,2 were significantly associated with attempted or completed suicide. Suicide in these instances may be viewed as a mean of escape from suffering.38

The finding that psychiatric medication prescriptions and hospitalizations for mental disorders (characteristic of the Increasing group) are related to completed suicide is consistent with Rahman et al.,18 who found that combined prescription of antidepressants with anxiolytics and inpatient care due to mental diagnoses or suicide attempt were strongly associated with completed suicide. Likewise, the high number of contacts with mental health services characteristic of the Increasing and Decreasing groups aligns with Denneson et al.,39 who found that 48% of the veterans had contact with mental health services in the year before death.

Limitations and strengths

The current study has some limitations. It uses secondary data (existing databases), preventing the analysis of variables that may be socially (e.g., living alone) or clinically relevant (e.g., diagnoses made at other health services, family history of suicide, or hopelessness), as well as selection of specific time frames for data, lack of confounder information, or unknown data quality. Another possible limitation is the accuracy of diagnosis or causes of death. However, it also has significant strengths. To our knowledge, it is the first investigation to shed light on a broad clinical profile of people who have died by suicide in Galicia (and one of the few existing studies of this nature, in Spain or internationally). It is also the first study in any location, to our knowledge, to address hospitalization trajectories prior to suicide and related variables.

Implications for research and clinical practice

Future research should include longitudinal designs and compare findings for people who die by suicide versus those who do not, allowing us to establish the causal factors for completed suicides. Data on other important clinical variables (e.g., suicidal ideation, hopelessness, impulsivity), as well as on diagnoses made and medications prescribed through other health services (mainly outpatient services) are recommended for inclusion in future research. In any case, this study's findings have important clinical implications. Establishing a profile for people who die by suicide in Galicia, as well as identifying the population subgroups with the highest rates (older men from rural areas), provide answers as to where and who, allowing clinicians to target individuals at-risk for death by suicide and prevention measures to those regions and populations most at risk. The question of when is illuminated by the fact that a high percentage of people who die by suicide have previously had contact with public health services and that two groups of hospitalization trajectories were identified (Increasing and Decreasing) representing 5% of the sample and who had both a high number of hospitalizations and frequent contact with other health services in the years prior to suicide. These insights mean that healthcare professionals may have opportunities for effective detection and prevention.

Conclusions

This investigation sheds light on the profile of people who committed suicide in Galicia between 2013 and 2016; it consisted of a retired man, with a mean age of almost 60 years, from urban and inner areas. The average suicide rate for that period was 12.3/100,000, the highest rates corresponding to men, aged 75–84, from rural areas. More than two thirds had had contact with public health services within the years prior to suicide. Almost half had been hospitalized, diagnosed with physical disorders and prescribed psychiatric medications; and almost one third had been diagnosed with mental disorders and received outpatient psychiatric care. Three trajectories of hospitalizations within the years prior to suicide were identified: a low number of hospitalizations, an increasing trend, and a decreasing trend. The increasing and decreasing groups had more psychiatric appointments and a higher incidence of mental and physical disorders. The strengths of this work are that it constitutes the first investigation to provide a broad clinical profile of people who committed suicide in Galicia and one of the few existing studies of this nature worldwide. These findings lay the foundation for the development of effective measures for the detection and prevention of suicide implemented by healthcare professionals from the public health service, and targeted at the most vulnerable populations.

Funding

This study was funded by the Galician Healthcare Service (SERGAS).

Conflict of interest

None.

Acknowledgements

We would like to thank the Galician Healthcare Service (SERGAS) for its support.

References
[1]
World Health Organization [WHO].
Mental Health Atlas 2017.
WHO, (2018),
[2]
Pan American Health Organization [PAHO].
Prevención del suicidio: un imperativo global.
PAHO, (2014),
[3]
P. Fernández-Navarro, M.L. Barrigón, J. Lopez-Castroman, et al.
Suicide mortality trends in Galicia Spain and their relationship with economic indicators.
Epidemiol Psychiatr Sci, 25 (2016), pp. 475-484
[4]
A. Álvaro-Meca, T. Kneib, R. Gil-Prieto, A. Gil de Miguel.
Epidemiology of suicide in Spain, 1981–2008: a spatiotemporal analysis.
Public Health, 127 (2013), pp. 380-385
[5]
I. Ruiz-Pérez, M. Rodríguez-Barranco, A. Rojas-García, O. Mendoza-García.
Economic crisis and suicides in Spain Socio-demographic and regional variability.
Eur J Heal Econ, 18 (2017), pp. 313-320
[6]
A. Shelef, J. Hiss, G. Cherkashin, et al.
Psychosocial and medical aspects of older suicide completers in Israel: a 10-year survey.
Int J Geriatr Psychiatry, 29 (2014), pp. 846-851
[7]
G. Cheung, S. Merry, F. Sundram.
Medical examiner and coroner reports: uses and limitations in the epidemiology and prevention of late-life suicide.
Int J Geriatr Psychiatry, 30 (2015), pp. 781-792
[8]
J. Nie, A. O’Neil, B. Liao, C. Lu, D. Aune, Y. Wang.
Risk factors for completed suicide in the general population: a prospective cohort study of 242,952 people.
J Affect Disord, 282 (2021), pp. 707-711
[9]
A. Clapperton, L. Bugeja, S. Newstead, J. Pirkis.
Identifying typologies of persons who died by suicide: characterizing suicide in Victoria, Australia.
Arch Suicide Res, 24 (2020), pp. 18-33
[10]
L. Jacob, H. Oh, A. Koyanagi, L. Smith, K. Kostev.
Relationship between physical conditions and attempted or completed suicide in more than 9,300 individuals from the United Kingdom: a case-control study.
J Affect Disord, 274 (2020), pp. 457-463
[11]
A. Paraschakis, I. Michopoulos, C. Christodoulou, F. Koutsaftis, L. Lykouras, A. Douzenis.
A 2-year psychological autopsy study of completed suicides in the Athens greater area, Greece.
Psychiatry Investig, 12 (2015), pp. 212-217
[12]
L. Lehmann, R.A. McCormick, L. McCracken.
Suicidal behavior among patients in the VA health care system.
Psychiatr Serv, 46 (1995), pp. 1069-1071
[13]
A. Erlangsen, W. Vach, B. Jeune.
The effect of hospitalization with medical illnesses on the suicide risk in the oldest old: a population-based register study.
J Am Geriatr Soc, 53 (2005), pp. 771-776
[14]
C. Basham, L.M. Denneson, L. Millet, X. Shen, J. Duckart, S.K. Dobscha.
Characteristics and VA health care utilization of U.S. Veterans who completed suicide in Oregon between 2000 and 2005.
Suic Life-Threat Behav, 41 (2011), pp. 287-296
[15]
A. Paraschakis, I. Michopoulos, C. Christodoulou, F. Koutsaftis, A. Douzenis.
Characteristics of suicide victims who had verbally communicated suicidal feelings to their family members.
Psychiatr Danub, 27 (2015), pp. 230-235
[16]
M. Innamorati, M. Pompili, V. Masotti, et al.
Completed versus attempted suicide in psychiatric patients: a psychological autopsy study.
J Psychiatr Pract, 14 (2008), pp. 216-224
[17]
G. Castelpietra, M. Bovenzi, E. Clagnan, F. Barbone, M. Balestrieri, G. Isacsson.
Diagnoses and prescriptions of antidepressants in suicides: Register findings from the Friuli Venezia Giulia Region, Italy, 2002–2008.
Int J Psychiatry Clin Pract, 20 (2016), pp. 121-124
[18]
S. Rahman, K. Alexanderson, J. Jokinen, E. Mittendorfer-Rutz.
Risk factors for suicidal behaviour in individuals on disability pension due to common mental disorders – a nationwide register-based prospective cohort study in Sweden.
[19]
National Institute of Statistics [INE].
Cifras oficiales de población resultantes de la revisión del Padrón municipal a 1 de enero.
(2020),
[20]
World Health Organization [WHO].
ICD-10: International Statistical Classification of diseases and related health problems.
WHO, (1992),
[21]
Galician Institute of Statistics [IGE].
Padrón municipal de habitantes [Municipal Census].
(2019),
[22]
Pan American Health Organization.
Indicadores de salud. Aspectos conceptuales y operativos.
Organización Panamericana de la Salud, (2018),
[23]
O.B. Ahmad, C. Boschi-Pinto, A.D. Lopez, C.K.L. Murray, R. Lozano, M. Inoue.
Age standardization of rates: a new WHO standard. GPE Discussion Paper Series.
(2001),
[24]
B.L. Jones, D.S. Nagin, K. Roeder.
A SAS procedure based on mixture models for estimating developmental trajectories.
Sociol Methods Res, 29 (2001), pp. 374-393
[25]
H. Lennon, S. Kelly, M. Sperrin, et al.
Framework to construct and interpret latent class trajectory modelling.
BMJ Open, 8 (2018), pp. e020683
[26]
S.L. Klijn, M.P. Weijenberg, P. Lemmens, P.A. Van Den Brandt, V. Lima Passos.
Introducing the fit-criteria assessment plot – a visualisation tool to assist class enumeration in group-based trajectory modelling.
Stat Methods Med Res, 26 (2017), pp. 2424-2436
[27]
C. Proust-Lima, V. Philipps, B. Liquet.
Estimation of extended mixed models using latent classes and latent processes: The R package lcmm.
J Stat Softw, 78 (2017), pp. 1-56
[28]
H. Lennon.
LCTMtools: latent class trajectory models: tools for checking adequacy. R package version 0.1.3.
(2020),
[29]
W.N. Venables, B.D. Ripley.
Modern applied statistics with S. Fourth Edition.
Springer, (2002),
[30]
C. Björkenstam, K. Alexanderson, E. Björkenstam, C. Lindholm, E. Mittendorfer-Rutz.
Diagnosis-specific disability pension and risk of all-cause and cause-specific mortality – a cohort study of 4.9 million inhabitants in Sweden.
BMC Public Health, 14 (2014), pp. 1247
[31]
I.P. Uribe, H. Blasco-Fontecilla, G. García-Parés, et al.
Attempted and completed suicide: not what we expected?.
J Affect Disord, 150 (2013), pp. 840-846
[32]
J. Alonso, M.C. Angermeyer, S. Bernert, et al.
Psychotropic drug utilization in Europe: results from the European study of the epidemiology of mental disorders (ESEMeD) project.
Acta Psychiatr Scand, 109 (2004), pp. 55-64
[33]
National Institute of Statistics [INE].
Estadística de defunciones según la causa de muerte. Resultados.
(2021),
[34]
S. Bachmann.
Epidemiology of suicide and the psychiatric perspective.
Int J Environ Res Public Health, 15 (2018),
[35]
S. Liu, A. Page, P. Yin, et al.
Spatiotemporal variation and social determinants of suicide in China, 2006–2012: findings from a nationally representative mortality surveillance system.
Psychol Med, 45 (2015), pp. 3259-3268
[36]
A. Sankaranarayanan, G. Carter, T. Lewin.
Rural–urban differences in suicide rates for current patients of a Public Mental Health Service in Australia.
Suic Life-Threat Behav, 40 (2010), pp. 376-382
[37]
K. Hawton, K. van Heeringen.
Suicide.
Lancet, 373 (2009), pp. 1372-1381
[38]
S.-T. Yeh, Y.-Y. Ng, S.-C. Wu.
Association of psychiatric and physical illnesses with suicide in older adults in Taiwan.
J Affect Disord, 264 (2020), pp. 425-429
[39]
L.M. Denneson, C. Basham, K.C. Dickinson, et al.
Suicide risk assessment and content of VA health care contacts before suicide completion by veterans in Oregon.
Psychiatr Serv, 61 (2010), pp. 1192-1197
Copyright © 2021. Sociedad Española de Psiquiatría y Salud Mental (SEPSM)
Descargar PDF
Opciones de artículo
es en pt

¿Es usted profesional sanitario apto para prescribir o dispensar medicamentos?

Are you a health professional able to prescribe or dispense drugs?

Você é um profissional de saúde habilitado a prescrever ou dispensar medicamentos