The negative consequences of the COVID-19 pandemic include physiological, psychological, social and economic burdens [see also our prior research on alcohol use disorder (6), as well as a perspective based on a small sample of patients in treatment for drug use disorder (7)]. These far-ranging effects might be particularly demanding for vulnerable groups, like patients suffering from substance use disorders (SUDs) (8). Serious implications for this subgroup including long-term socioeconomic and public health effects can be anticipated (9). In particular, increased risk of infection and severity of COVID-19 symptoms, psychological stress and reduced access to addiction treatment services are of major concern.

From a physiological perspective, substance use disorders (SUDs) were found to increase the risk to contract COVID-19 (10). Persons suffering from drug use disorder often develop conditions regarding the respiratory system from inhalation drugs. An overall impaired immune system as well as damaging effects of drug use on the cardiovascular system further increase the risk of mortality associated with COVID-19 (11). As patients suffering from SUDs are at higher risk for COVID-19 and worse outcomes, individual worries about the physiological effects of the pandemic could be anticipated.

Demanding psychological aspects of the pandemic and lockdowns are evident. Major psychological stressors are driven by trauma exposure, levels of perceived stress and isolation, rendering risk factors for a deterioration of symptoms of depression and anxiety (12). An Italian study investigating psychopathological burden during the beginning of the pandemic found relatively high rates of depression, anxiety, irritability, and post-traumatic stress symptoms among patients with SUDs (13).

Negative economic effects are clearly anticipated, since global economy is struggling heavily with the financial consequences of the pandemic. Loss of income due to reduced working hours, or even job loss due to the pandemic represent major economic stressors on the individual level, and might be a source for further psychological burden (14). Lower perceived economic stability additionally promoted the risk of post-traumatic stress symptoms during the pandemic (15). Income reduction further elevates the risk for depression and anxiety (12).

Social interactions have been heavily restricted during the pandemic due to lockdowns and other government measures. In Austria, social life was interrupted by closure of bars and restaurants, and a ban on large public gatherings. Even social interactions in private parts of life had to be immensely reduced, and restrictions on non-essential movement (except medical care, food shopping, or exercise) further promoted isolation during the second wave of the pandemic. Taking this situation into account, a tremendous burden on patients with SUD stemming from reduced social support as a protective factor (16) could be expected. Since substance use often occurs in social contexts, a decrease in consumption for recreational users might be observed during lockdowns. However, regular substance use and more severe SUDs might probably increase (12).

Already before the pandemic, high rates of post-traumatic stress disorder (PTSD) among patients with SUDs, but a low detection rate in treatment settings was assumed (17). In general, PTSD follows traumatic events and is characterized by a symptom pattern of intrusions, avoidance of thoughts and behaviors, negative changes in thoughts and mood, and changes in arousal and reactivity (18). Prior clinical research also confirmed relatively high rates of comorbid affective and anxiety disorders among patients in treatment for SUDs—a subgroup, which might also be characterized by a higher severity of this disorder (19). Furthermore, a complex interplay between psychiatric comorbidities and substance use is assumed. Among patients with opioid use disorder (OUD) depression has been identified as highly prevalent, and its impact on therapy outcome is anticipated, but poorly understood yet [for a recent review see Ghabrash et al. (20)]. The potential interplay between stress and risk for drug use was investigated among a sample of patients with OUD (21). Higher reported levels of stress have already been associated with early drop-out (22).

A rise in PTSD, anxiety and depression symptoms during the pandemic have been anticipated and confirmed in the general population (3, 23, 24). PTSD due to COVID-19 was expected as a common psychiatric response to the current pandemic and its related psychological stressors (25). Studies conducted in China and Italy during the initial stage of the pandemic, which were heavily affected areas, reported high rates of PTSD and psychological distress in the general population (26, 27). For patients with SUDs during this ongoing pandemic, negative mental health consequences including higher levels of depression, anxiety, irritability, and post-traumatic stress symptoms have already been confirmed (10, 12, 26). The COVID-19 pandemic renders an additional major source of distress for patients in opioid substitution therapy (OST), who are already more vulnerable in respect to their mental and social health. Close monitoring of this subgroup and providing stable OST services for this population is therefore mandatory during this crisis (28).

Misuse of the OST medication (29) and concomitant use of other illicit drugs is highly prevalent, and therefore a major issue of concern in OST. A Swiss registry-based study, which was conducted before the pandemic (30) found that more than a third of all participants reported at least one cocaine consumption day in the past month. Furthermore, a positive association between the dosage of methadone and concomitant use of cocaine was observed. Australian patients receiving OST had a significant reduction in the depression subscale of the DASS-21 after 3 months of treatment, less pronounced improvements were seen in the stress and anxiety subscales (31). Compared to normative values patients in methadone maintenance treatment had higher stress, post-traumatic stress symptoms and cortisol levels (32). Data from an US-study showed, that patients, dropping out from OST, reported higher levels of stress, heroin- and cocaine-craving than participants, who stayed in OST during the observation time (22).

In the context of the pandemic, in our previous study (N = 32) 79% of the participants in OST indicated concomitant use of other illicit drugs during the initial phase of the pandemic (7). However, this number has to be interpreted with caution, given potential biases due to the small sample size and a high proportion of inpatient participants in this study. Developments on illicit drug markets due to the pandemic, as well as their direct and indirect consequences remain unclear. Due to government control strategies and border closures major interruptions in illegal drug supply were expected (33). Unavailability of substances could lead to hazardous activities, including self-manufacturing of substances or even a rise in cases of suicide (34). Increase in pricing and disruption of illicit opioids could have further severe impacts on the drug-taking community, including more cases of overdose (11). This risk is heightened by the consumption of other opioids than normally administered due to the lack of availability, as well as by accompanied changes in quality and strength of those substances (28, 35). Furthermore, social distancing may increase the probability of fatal overdoses due to isolation without opportunity for rescue (36, 37). Consequently, the situation on the drug market should be closely monitored, enabling reactions to further potentially negative implications for patients suffering from drug use disorder.

Original data on patients suffering from drug use disorder, including those in OST during this ongoing pandemic are still sparse. Taking findings of studies focusing on SUDs in general (13) into account, an elevated risk to develop PTSD symptoms as a result of the crisis might be expected, and has to be monitored in this vulnerable group. Therefore, the main goal of the current study was to assess the presence and severity of PTSD symptoms due to the COVID-19 pandemic. To that end, PTSD symptoms were evaluated using an adapted version of the IES-R (38). The sample of patients in OST was accordingly split into two subgroups labeled as low or high risk for PTSD due to the pandemic based on the IES-R (but not as a professional diagnosis of PTSD). In this context, the impact of potentially contributing sociodemographic and various COVID-19 related worries and fears for different areas of life (physiological, psychological, economic and social factors) were investigated. Furthermore, levels of severity in psychopathology (depression, anxiety, and stress), as well as differences and changes on these measures since the beginning of the pandemic were evaluated between the two groups. Additionally, momentary craving, concomitant use of illicit substances, and developments on the Austrian drug market were assessed.

For this cross-sectional survey study, data was collected from patients receiving treatment at two outpatient facilities in Austria. The duration of the study was 14 weeks, between end of November 2020 and beginning of March 2021. Only patients, who were currently in OST, and provided responses on nearly all of the items of the survey (defined as a maximum of four missing responses on the scales) were included in the final analysis, resulting in a total sample of N = 123. This study was conducted in accordance with the Declaration of Helsinki and approved by the local ethics committee. Participants provided written informed consent, and data was processed and analyzed anonymously. Data collection started after a new increase of COVID-19 incidence in Austria—also called the second wave—between December 2020 and February 2021. During this time period, hotels, restaurants, and bars remained closed, and social interactions were restricted in public and private areas of life by government measures.

Data was analyzed using IBM SPSS Statistics for Windows (Version 25.0) (46). Descriptive statistics of the variables are reported in Table 1. The IES-R was adapted to assess PTSD symptoms exclusively for the COVID-19 pandemic (and no other traumatic events). Main analyses of this study were based on the cutoff score for being at risk of PTSD according to the IES-R total score. To that end, the total sample was split into two subgroups of patients indicating low or high risk of PTSD due to the pandemic [for more details see section Impact of Event Scale (IES-R) Adapted for COVID-19 below].

To evaluate the potential impact of sociodemographic variables and different COVID-19 factors on the risk of PTSD due to the pandemic, a binary logistic regression analysis was conducted. Furthermore, differences between the low- (N = 90) and high-risk (N = 33) PTSD-groups were evaluated for clinical symptoms (depression, anxiety, and stress) as well as for craving, using Mann Whitney tests. Changes on the symptomatology were assessed, and differences between the groups were further investigated. Findings on concomitant use of other illicit substances, and developments on the Austrian drug market are reported in a descriptive manner. Effect sizes for the different analyses are reported as correlation coefficient r and interpreted according to Cohen (47) as small (0.1–0.3), moderate (0.3–0.5), and strong (>0.5) effects.

Descriptive statistics for the IES-R, sociodemographic variables, and COVID-19 factors are displayed for the total sample and the two subgroups (low- and high-risk PTSD) in Table 1.

The IES-R (38) was adapted to evaluated PTSD symptoms due to COVID-19. In this sample, excellent internal consistency was found for the total IES-R score (Cronbach’s α = 0.91), and moderate to high levels for the three subscale scores (Intrusion α = 0.78; Avoidance α = 0.82; Hyperarousal α = 0.76). Rash et al. (42) examined a range of cutoff scores for the IES-R for suitability with a substance dependent sample. Their results indicated a recommended cutoff score of 22–24 on this scale to determine an elevated risk of PTSD. Cutoff values of 22–24 in this study met the goal to maximize sensitivity (92%, specificity of 57%), with an overall correct classification rate of PTSD cases of 77%. Based on these findings, a cutoff value of 24 was selected for this study. Accordingly, the total sample was split into two subgroups of patients indicating low (i.e., total IES-R score <24) or high risk of PTSD (i.e., total score ≥24) due to the pandemic. As confirmed by Mann-Whitney tests, the two subgroups did not only differ significantly in the total score of the IES-R (z = −8.5, p < 0.001), but also on all three subscales for intrusion (z = −7.9), avoidance (z = −7.6), and hyperarousal (z = −7.3, all ps < 0.001, all rs > 0.65; see descriptive data in Table 1).

A binary variable was constructed for patients at low-risk (value = 0; N = 90) or high-risk (value = 1; N = 33) for PTSD due to the COVID-19 pandemic according to IES-R scores. A binary logistic regression analysis was then performed to investigate potential risk factors for PTSD. The model allows to evaluate the effects of sociodemographic factors (age, gender, living alone, and employment) and COVID-19 impact (physiological, psychological, economic, and social factors) on the probability of experiencing PTSD symptoms due to COVID-19. A backward variable selection procedure (Wald) was performed using a cutoff value of 0.27 (i.e., the proportion of patients with high-risk for PTSD in the total sample). Results of this regression analysis are presented in Table 2, for the initial model as well as for the final model after variable selection.

The final model (step 8 with a correct classification rate of 0.71) included psychological and economic COVID-19 factors as predictors, and was statistically significant, χ²(2) = 17.1, p < 0.001. Nagelkerke R² of 18.9% showed a moderate goodness of fit of the model with moderate to high levels of sensitivity (0.64) and specificity (0.73). Patients indicating a stronger negative impact by psychological COVID-19 factors had a higher risk (odds ratio of 1.21, p = 0.007) for PTSD. Economic COVID-19 factors (odds ratio of 1.14, p = 0.045) also increased the probability for PTSD according to IES-R scores.

The DASS was originally constructed to measure self-reported negative emotional states of depression, anxiety and stress, including 42 items (48). The short version DASS-21 (44) consists of 21 items (ranging from 0 to 3) with seven items per subscale. In the current study, the sum of all item scores was calculated for the total score (ranging from 0 to 63). For the subscores of depression, anxiety and stress the item scores of the respective subscales were summed, respectively¹. For this measure, levels of severity and respective cutoff scores for the subscales were adapted from the original DASS (48). High internal consistency was found for the DASS-21 total score (Cronbach’s α = 0.95), and the three subscale scores (Depression α = 0.92; Anxiety α = 0.82; Stress α = 0.89).

In our total sample of patients in OST, 48% reported concomitant use of non-prescribed illicit substances. A qui-square test of independence was conducted to evaluate potential associations between risk for PTSD (low- vs. high-risk) and concomitant use (no vs. yes). Results confirmed a significant association between the two variables, X²(1, N = 123) = 4.4, p < 0.05, r = 0.19. This finding suggests that those reporting concomitant use were also more likely to be part of the high-risk PTSD group.

Among this group indicating concomitant use (N = 59, see Table 4 for descriptive statistics), most reported consuming cannabis (80%), followed by heroin (24%), cocaine (17%), and unprescribed benzodiazepines (15%). Only a small proportion indicated consumption of methamphetamines (3%), hallucinogens (3%), or amphetamines (2%). In respect to legal substances, 17% indicated drinking alcohol regularly and 86% in this group were smokers. The subgroup of patients reporting concomitant use illicit drugs had a mean score of 9.23 (SD = 3.1) on the DUDIT-C, and indicated no relevant changes in frequency (mean = −0.1) and quantity of consumption (mean = −0.3) Developments on the Austrian drug market were evaluated for pricing, availability problems and quality of illegally purchased substances (on scales from −5 to +5). Although about 15% of these patients reported an increase in both, prices and availability problems, the majority indicated no change on the three scales, suggesting a rather stable situation on the Austrian drug market. In sum, no substantial changes could be detected based on the patients’ responses, with only slight increases in prices (mean = 0.2) and availability problems (mean = 0.8), and a mean decrease in quality (mean = −0.7).

One third (N = 41) of patients in our total sample had an additional prescription of benzodiazepines. A qui-square test of independence did not result in a significant association between risk for PTSD (low- vs. high-risk) and benzodiazepine prescription (no vs. yes), X²(1, N = 123) = 0.2, p = 0.67, r = 0.04. This finding suggests, that patients with an additional prescription of benzodiazepine were not at higher risk for PTSD due to the pandemic.