Suicides in India: A Forecasting for Immediate Investment to Save Lives

Abstract

Background:

Suicide, a preventable death, is a global public health problem. As per Centers for Disease Control and Prevention, there is 1 death every 11 minutes. Seventy-seven percent of global suicides occur in low- and middle-income countries. In India, there were 12.4 suicides per lakh population, with 170924 suicides in the year 2022. To plan for an effective intervention, a predictive analysis i.e. forecasting the future risk of suicides, is more important than exploratory data analysis.

Methodology:

The National Crime Records Bureau, GoI, publishes annual reports every year on accidental deaths and suicides in India. Data on suicides were extracted from these reports (1967–2021). Data analysis was done using Gretl Software (2023a) with ARIMA modeling, and suicide cases were forecasted for the following 5 years (2022–2026).

Results:

There has been a gradual increase in the number of suicides in India, which started declining in the year 2012 until 2017, and then again started increasing, with a drastic increase in the number of suicides in the year 2020 during the Covid-19 pandemic. Forecasting of suicides by ARIMA modeling for the next 5 years from 2022 to 2026 shows an increasing trend, with a gradual decrease in the rate of increase in the absolute number of suicides.

Conclusion:

There is an increase in the absolute number of suicides in India. Though the rate of increase in suicides every year is going down, the absolute number of suicides is still a public health concern. This gives a warning sign for an upcoming epidemic in India; hence, it is time to invest in it to prevent these suicide events.

BACKGROUND

Suicide, a preventable death, is a global public health concern. According to data from the Centers for Disease Control and Prevention (CDC), there is 1 death every 11 minutes. For each suicide death, there are 4 hospitalizations for suicide attempts, 8 emergency department visits, 27 self-reported suicide attempts, and 275 people who seriously considered suicide.[1] Hence, each suicide event is associated with many such hidden social occurrences. Seventy-seven percent of global suicides occur in low- and middle-income countries.[2] Suicide is one of the leading causes of death and a major public health problem in developing countries such as India. In India, there were 12.4 suicides per lakh population, with an absolute count of 170924 suicides in the year 2022.[3] Although the emphasis in the healthcare system has shifted from physical care to both physical and mental healthcare, the mental healthcare component is still not universally included in our day-to-day healthcare services.

Toward the end of 2019, the COVID-19 disease began spreading globally. The World Health Organization (WHO) declared it a Public Health Emergency of International Concern on January 30, 2020, and later declared as a pandemic on March 11, 2020.[4] The COVID-19 disease was a public health crisis and overwhelmed people’s daily lives globally.[5] COVID-19 broke the economic backbone of all countries around the globe and affected the socio-economic status of individuals in all countries, which indirectly impacted their mental status of individuals. According to studies, a rapid rise in unemployment due to the COVID-19 pandemic was predicted to result in 3,235–8,164 excess suicides between 2020 and 2021, representing a 3.3%–8.4% increase in suicides per year compared to the 2018 rate of 48,432 suicides in the USA.[6] It was also noted by McIntyre et. al. that an abrupt increase in unemployment in Canada related to the COVID-19 pandemic is associated with an increase in deaths due to suicide.[7] Similar studies are lacking in India, but we have all witnessed the effect of COVID-19 on employment and, indirectly, on the mental health status of individuals in India.

The above studies and reports highlight the burden of suicides, which is a preventable/avoidable event. From a public health perspective, it is likely to become a future epidemic in developing countries like India, where there is higher population growth and a lack of mental healthcare in day-to-day health services. To plan for an effective intervention, predictive analysis—i.e., forecasting the future risk of suicides—is more important than exploratory data analysis i.e., suicide pattern analysis.[8] There have been many suicide death prediction models, but COVID-19 has affected all such predictions. Furthermore, the various models to predict/forecast suicide deaths or rates have their own advantages and limitations.[9,10] Even though there are models to predict individual suicidal behavior or risk, in a country with a huge population such as India, this kind of behavioral prediction does not work. It is feasible only in health facilities where the course of treatment can be assessed. Furthermore, the accuracy of predicting suicidal behavior based on clinical judgment varies considerably across physicians, and risk factors known to be strongly associated with suicidal behaviors are weak predictors on their own.[11]. Still there are many models to predict suicidal behavior or the risk of patients, though they all have many limitations.[12] Though predictions/forecasting at a geographical level have limited clinical value, from a public health perspective, they can help policymakers focus and plan effective intervention to address the problem. A time series analysis of suicide death data in India before the COVID pandemic showed an observable and rising trend in suicide rates in India over the last 5 decades. The forecast indicates a continuance of rising suicide cases for the next couple of years in India, with a limited decline in the following years.[13] However, all these predictions are affected by the COVID-19 pandemic, which is a limitation of such forecasting. Keeping in mind the high burden of reported suicide events in India and the impact of the COVID-19 pandemic, a forecasting of the number of suicides in India was conducted.

METHODOLOGY

A time series analysis of the number of suicides in India was conducted, for which data related to the number of suicides per year were collected from sources available in the public domain. The National Crime Records Bureau, Ministry of Home Affairs, Government of India, publishes annual reports on Accidental Deaths and Suicides in India (ADSI). The published annual reports on accidental deaths and suicides were compiled from the NCRB website, and the data on suicides were extracted from the annual reports. By the time the study was conducted, data on suicides in India were available from 1967 to 2021. The extracted data were entered in a Microsoft Excel Sheet, and data analysis was performed using Gretl Software (2023a)[14] with auto regressive integrated moving average (ARIMA) modeling.[15] ARIMA modeling is one of the most commonly used tools for prediction/forecasting due to its ease of application and accuracy. The ARIMA model developed by Box and Jenkins consists of three steps: model identification, parameter estimation, and model validation. The requirement for ARIMA modeling is a time series data. Before processing the dataset for ARIMA modelling, it should be checked for stationarity. In the current study, prior to ARIMA modeling, a time series analysis was conducted to study the pattern of suicides in India from 1967 to 2021. Then, the dataset collected was checked for stationarity. Stationarity means that values vary over time around a constant mean and variance. Stationarity was checked using the autocorrelation function (ACF) and partial ACF (PACF) of the primary dataset. Nonstationary data were transformed into stationary data by applying appropriate differencing, and the order of differencing was used as one of the parameters for ARIMA model development (model identification). A prediction model was developed by considering the other components of the model i.e. auto regression and moving average. The parameters to be used for the ARIMA modelling, i.e. the auto regression, difference of order, and moving average, were estimated, and a robust model was developed (parameter estimation) based on the lowest Akaike information criterion. Once the model was ready, it was validated by predicting the number of suicides for the sample period (model validation), which was from 1967 to 2021. However, the predictions for the sample period were from 1969 to 2021 because the software used the previous years’ data to predict the number for the following years. After the validation, the model was used to predict/forecast the number of suicides for the following 5 years. By the time this manuscript was prepared, data on suicides for the year 2022 (ADSI 2022 Report)[3] was available and was kept to verify the post-sample prediction/forecasting. The prediction was presented with an absolute number of suicides year-wise, along with a 95% confidence interval. Institute Ethics Committee approval was obtained for the study.

RESULT

Data on suicides in India were available for 55 years, with the number of suicides ranging from 38829 to 164033 from the year 1967 to 2021, respectively. There was no significant pattern until 1980, and after which there was a gradual increase in the number of suicides in India. This trend started declining from 2012 to 2017 and then began increasing again, with a drastic rise in the number of suicides in 2020 during the COVID-19 pandemic. The increase in suicides in 2020 was more than 10% of the reported suicides in 2019, and the increase in the number of suicides in 2021 was 7% of that in 2020. This clearly shows the effect of COVID-19 on suicides in India due to various causes. As per the ADSI 2022 report, the number of suicides was 170924, which was 4% higher than the number reported in 2021. The trend shows an increase in the absolute number of suicides in India every year, but the absolute difference between the years is decreasing every year [Figure 1].

Figure 1.

Trend of Suicides in India from 1967 to 2021

As presented in Figure 2, the forecasted number of suicides per year almost overlaps with the actual reported suicides per year in the sample period, which represents the accuracy/validation of the forecasted ARIMA model. The post-sample period, or forecasted period from 2022 to 2026 (five years), shows an increasing trend of suicides in India with the same inclination/slope as observed from 2018 to 2021. In the post-sample period, the asterisk represent the forecasted absolute number of suicides, and the range represents the 95% confidence interval. The absolute number and 95% confidence interval for the post-sample 5 years are presented in Table 1. The table shows an increase in the confidence interval with an increase in the number of years forecasted. The closer the year, the better the forecasting, with a 95% confidence interval.

Figure 2.

Forecasting of Suicides in India with 95% CI from 2022 to 2026

Table 1.

Forecasted absolute numbers of Suicides and 95% CI in India from 2022 to 2026

Year	Number of Suicides	LL of 95% CI	UL of 95% CI
2022	176544	169408	183680
2023	188648	175379	201918
2024	201184	179954	222414
2025	213789	183658	243920
2026	226619	186533	266704

DISCUSSION

Data for the study were extracted from the ADSI reports published by NCRB, but it should be noted that these statistics are based on police reports; hence, the numbers are likely less than the actual number of events and are under-reported. Despite the under-reporting of suicides, the NCRB offers significant insight for the planning of suicide interventions based on the reports.[16] The trend analysis showed an increasing trend[17] of suicides in India, which drastically increased during the COVID-19 pandemic, similar to other countries. This reflects the effect of the COVID-19 pandemic on the mental health of individuals. The same pattern was also observed in other countries.[6,7] There have been previous forecasting studies on suicides, but the pandemic affected all these studies and rendered the forecasted trends incorrect. The rate of increase in the number of suicides after COVID-19 has started declining, but the absolute numbers are still on the rise, which is a national concern. The current study showed an increasing trend in the following 5 years. Although these findings are of no clinical importance, they will help policy makers shape better community-level interventions to prevent such events in India. Suicides are preventable, as we are aware that for each suicide, there are hundreds of associated morbidities, hospitalizations, and visits to hospitals, which increase the burden on our health system and the affected family. It was observed that the number of suicides is increasing with the rise in the total population in India. As India has now surpassed China to become the most populous country in the world, and the rate of suicides is also increasing, it can be imagined that, soon, this may take the shape of an epidemic in the country—an epidemic without any visible cause to address. Hence, this is the time to invest to halt the rise in suicides in India, as all deaths due to suicides are unnatural and entirely preventable. The life of the person can not only be saved, but the family can also be saved by saving a single life by preventing suicidal deaths. The increasing rate of suicides is not on par with the population growth rate, but, indeed, the number of suicides cannot be ignored. In most cases, it is the working group that commits suicide, and that affects the socioeconomic growth of the family and nation. By the time this paper was written, the ADSI 2022 report[3] was available, and it was noticed that the absolute number of suicides (170924) in 2022 was within the forecasted 95% CI (169408: 183680) for 2022 by the selected ARIMA model, which shows the power of the forecasting model.

The limitation of ARIMA modeling lies in its assumption that the pattern of event occurrences remains consistent over time, which is attributed to the stationarity of the dataset. A sudden shift in the pattern of event occurrences, driven by various factors, can significantly affect the accuracy of predictions or forecasts. One such example is the COVID-19 pandemic, an emerging disease with unknown epidemiology, which significantly impacted the occurrence of many health events for various reasons. As a result, any predictions or forecasts of health events made prior to the COVID-19 pandemic were rendered inaccurate. However, the strength of ARIMA modeling lies in its simplicity and accuracy, which is why it remains one of the most commonly used predictive models in research and surveillance.

CONCLUSION

The study finding shows that there is an increasing trend of suicides in India, and the forecasting model has shown that the increasing trend will continue in the next 5 years. Though the rate of increase in suicides every year is going down, still the absolute number of suicides is a public health concern. This gives a warning sign for an upcoming epidemic in India; hence, it is time to invest in mental health education and services to deal with such conditions to prevent these events. We always say “prevention is better than cure,” and for events like suicides, there is no cure—it is only prevention that must be considered.

Key messages

There is an increasing trend of suicides in India that will continue in the same trend for the next 5 years. This is a warning sign for an upcoming epidemic in India. Now is the time to invest in halting the epidemic.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.