An interrupted time series analysis of hospital admissions due to alcohol intoxication during the COVID‐19 pandemic in Tehran, Iran

The COVID-19 outbreak affected mental health globally. One of the major concerns following the COVID-19 pandemic was increased incidence of risky behaviors including alcohol consumption. This study evaluates the trend of alcohol poisoning in Loghman-Hakim Hospital (LHH), the main referral center of poisoning in Tehran, during the 2-year period from 1 year prior to 1 year after the onset (February 23rd, 2020) of the COVID-19 epidemic in Iran. All patients admitted with alcohol intoxication from February 23rd, 2019 to February 22nd, 2021 were evaluated and patient data extracted from LHH electronic hospital records. Alcohols were categorized as toxic (methyl alcohol) and non-toxic (ethyl alcohol). Of 2483 patients admitted, 796/14,493 (5.49%) and 1687/13,883 (12.15%) had been hospitalized before and after the onset of the COVID-19 epidemic in Iran, respectively. In total, 140 patients did not survive, of whom 131 (93.6%) were confirmed to have methanol intoxication. Mortality was significantly higher during the outbreak (127 vs 13; P < 0.001; OR: 4.90; CI 95%: 2.75 to 8.73). Among the patients, 503 were younger than age 20. Trend of alcohol intoxication showed increases in children (57 vs 17) and adolescents (246 vs 183) when compared before and after the COVID-19 epidemic outbreak. A total of 955 patients were diagnosed with methanol toxicity which occurred more frequently during the COVID-19 era (877 vs 78; P < 0.001; OR: 10.00; CI 95%: 7.75 to12.82). Interrupted time series analysis (April 2016–February 2021) showed that in the first month of the COVID-19 epidemic (March 2020), there was a significant increase in the alcohol intoxication rate by 13.76% (P < 0.02, CI = [2.42–24.91]). The trend of alcohol intoxication as well as resulting mortality increased in all age groups during the COVID-19 epidemic in Iran, indicating urgent need for the prevention of high-risk alcohol use as well as improved treatment.

Age categorization. Patients aged 12 and younger were considered as children. The adolescent group consisted of the patients 13-20 years of age and all those 21 years and older were considered as adults.
Type of alcohol. All patients with serum methanol level > 6.25 mmoL/L (20 mg/dL) were considered as methanol-intoxicated (either solely or in a co-ingestion with other medications/drugs/substances) 27 . Patients with an ethanol level higher than 50 mg/dL were considered as ethanol-poisoned again either solely or in combination with other medications/drugs/substances 28 . Epidemiological data and COVID-19 categorization. Patients' demographic characteristics (age, gender), final diagnosis, and their final outcome were retrieved from the LHH electronic archive and analyzed.
Missing data and statistical analysis. Data was analyzed using statistical package for social sciences (SPSS) software version 26 (IBM CO., Chicago, Ill, USA). Comparing parts with missing and non-missing data, it appears that missing data was completely random and randomly distributed across all observations. Chi-Square and Fisher's exact tests were applied to search for significant differences in the categorical variables. The 95% confidence interval (CI) was used to estimate the precision of the odds ratio (OR). Mann-Whitney test applied to see differences among age and expressed as median [Inter quartile range]. Stata 14.0 (Stata Corp LLC, College Station, Texas, USA) was used for interrupted time series analysis to see whether the COVID-19 pandemic resulted in a shift in the level and trend of alcohol intoxication compared with those of the pre-COVID-19 period. We set the significance level at alpha = 0.05. For specific aim of time series analysis, data on the trend of alcohol intoxication was also gathered from April 2016 till February 2019 for a better assessment of pre-pandemic trend. Alcohol type and mortality. In the 2-years scope of the study, 140 (5.64% of total alcohol cases and 16.77% of total mortalities due to poisoning) alcohol intoxicated patients died (21 females and 119 males). Mortality rates were 1.6% and 7.5% of alcohol intoxicated cases before and after the COVID-19 epidemic, respectively (13 versus 127 deaths; P < 0.001; OR = 4.90; CI 95% = 2.75-8.73). Figure 1 shows the trend of alcohol intoxication (Fig. 1A) and mortality (Fig. 1B)  www.nature.com/scientificreports/ significantly higher among methanol-poisoned patients (P < 0.001; OR = 6.54; CI 95% = 4.35-9.84). Of 140 deaths, 131 (93.6%) had occurred among the patients diagnosed with methanol poisoning (mortality rate = 11.75%). Of this note, 5 patients (3.6%) consumed methanol in addition to at least one substance/medication. Almost 27% (665 patients) of our cases had been admitted with mixed poisoning of alcohols and other substances/medications, of whom 44 and 621 were finally diagnosed with methanol and ethanol intoxication as the alcohol involved in the intoxication, again with no significant difference between the pre-and during-COVID epidemic. Table 2 shows the pair comparisons of three groups together. Mortality was significantly higher before and during-COVID epidemic while methanol was compared to ethanol (P = 0.016; OR = 29.41; CI 95% = 2.41-333.3) or to the patients who had co-ingestion (P = 0.045; OR = 9.71; CI 95% = 7. 25-13.18). No significant differences were found between ethanol and co-ingestion group in alive or dead cases any time. Only 11 patients with mixed poisoning did not survive (mortality rate = 1.65%). Figure 1C,D depict the trend of poisoning with methanol and mixed poisoning during the study period, respectively. Trend of alcohol intoxication among children (A), adolescents (B), and Adults (C) Pre-COVID 19 and during COVID-19 are shown in Fig. 2. Table 3 and Fig. 3, the alcohol intoxication rate increased every month before March 2020 by 0.04%. This increase was statistically significant (P = 0.03, CI = [0.01-0.07]). www.nature.com/scientificreports/  The trend of alcohol intoxication after the COVID-19 pandemic (relative to the pre-COVID-19 trend) decreased by 1.44% per month, which was not statistically significant (P = 0.05, CI = [− 2.91 to 0.02]).

Discussion
As far as we know, this study reports the highest number of methanol poisoning cases in a single center in the world addressing the trend of alcohol intoxication before and after COVID-19 pandemic 29,30 . Based on our results, a rough twofold increase is detected in the number of alcohol poisoned cases during the pandemic. In addition, there was a discernible rise in the first 3 months after the beginning of the outbreak. During the pandemic, people experienced numerous adversities in their lives including lockdowns and social distancing, distant working, unemployment, loss of the family, and school closures 31 . During the early days of the pandemic, it was challenging for many people to come up with a variety of stressful stimuli including the fear of the disease and uncertainty about the future 32 . Studies have reported increased anxiety and distress in several countries 33,34 .
Increasing alcohol consumption was a coping behavior to this elevated anxiety 32 . Evidence shows dramatic increase in the alcoholic beverages sales 35,36 even though several countries utilized restrictive measures to control alcohol consumption 24 .
Working from home needs adjustment in daily routines and consequently causes an elevated anxiety state 34 . Unemployment acted similarly and resulted in higher alcohol consumptions 32 . Evidence from former economic crises also highlights similar consequences 37 . In such situation, despite being unemployed, people may tend to use cheap alcoholic beverages in a cost-saving manner 38 . Solitary drinking is also believed to be a risk factor for deleterious outcomes of alcohol consumption 39 .
Additionally, patients with COVID-19 infection and their families as well as the health care workers experience stigmatizations resulting in stressful stimuli 40 . There are reports of concomitant COVID-19 infection and methanol poisoning 41 . Loss of friends and relatives and complicated grief are of the other factors altering alcohol consumption behavior 42 . Having to provide childcare while working from home also reported to be stressful for the parents during the pandemic 29,31 . Lockdown and restrictive measures have also made school and university student another vulnerable group; it shown that interactions with teachers and classmates have a fundamental role in reducing mental health problems including anxiety and depression 43 .
The pandemic has also been associated with an increase in depression in the public 44 . Many people with psychological and mental disorders may be deprived from obtaining appropriate care due to the fear of becoming infected or reduced resources due to restrictive measures 31 .
In crises, people seek for new information to be up to date from any source possible which may predispose the society to the danger of false information and rumors 2 . Based on our results, during the first 3 months of the outbreak, a discernible rise was spotted in the number of alcohol-poisoned cases and specifically in the numbers of methanol-poisoned patients. It was erroneously propagated that alcohol prevented COVID-19 infection. People started to drink alcoholic beverages and even sanitizers to prevent infection. On the other hand, increased demand for hand sanitizers led to scarcity and finally indulgence of the market with low-quality products containing methanol 45 .
The increased risk of poisoning was detected in all age groups. It was initially believed that alcohol poisoning was more suicidal or deliberate among adults and adolescents and more accidental among children 46 . However, www.nature.com/scientificreports/ our results showed that even in the adults it was more recreational. This could also be due to stocking of lowquality alcohol-containing products and sanitizers at homes. In Iran, alcohol drinking is prohibited, and the black market provides alcoholic beverages to the public. There is no national center for monitoring alcohol consumption in the society, and data on its sale and use are missing. However, a recent study suggests the rising trend of alcohol poisoning through the recent years 47 . Outbreaks of methanol poisoning in Iran have been occurred frequently and have been always a public health concern 48-50 . Our data suggests a sharp rise in methanol poisoning in the first 3 months after the outbreak resulting in high mortality in the total population and specifically among adolescents. Of seven deceased adolescents, methanol poisoning was diagnosed in six, all happening in those 3 months. Another fact is that methanol poisoning regularly peaked during the year preceding the COVID-19 epidemic but after that it never reached a zero level. This is another concern implying the fact that the market is continuously indulged with low-quality products resulting in ongoing methanol poisoning.
Due to the prohibition of alcohol consumption and legal consequences, intoxicated people usually avoid referring to the health care facilities until the final stages of the poisoning resulting in a higher mortality and morbidity 51 . This emphasizes the importance of active case finding 52 . Limitations. Considering the retrospective nature of the study, we could not retrieve the intent of alcohol ingestion in many cases; also, we were not able to follow the patients after discharge to check for possible mortality post-discharge. The generalizability of this single-center study is unclear. The current study may be subject to selection bias, as we did not measure methanol and ethanol levels of all poisoned patients. Measurement was done only in those with a history of alcohol ingestion or clinical finding/lab exams in favor of alcohol intoxication.

Conclusion
Market surveillance and increasing public awareness should be intensified to prevent further methanol poisonings during the COVID-19 epidemic in Iran. Public education on dangerous effects of non-standard alcoholic beverages and sanitizers as well as proper stocking of these products at home can decrease alcohol poisonings in all age groups.
Measures to control alcohol consumption and suicidal attempts should be followed more intensely among adolescents and social support should be provided for this age group especially during crises of this kind as more risky behaviors are anticipated in them 53 . This is extremely important as their drinking habit in this period may affect their drinking habit as an adult 54,55 .
It should be borne in mind that monitoring and measures should not be limited to the pandemic era. Previous experiences from the SARS epidemic 55

Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.