Contact with primary care physicians among adults with pre-existing common mental health problems during the COVID-19 pandemic: a registry-based study from Norway

On March 11, 2020, the coronavirus disease 2019 (COVID-19) was declared a pandemic by the World Health Organization [1]. The measures to prevent the spread have not only had a considerable impact on physical and mental health outcomes [2], but also affected the delivery of mental health care [3]. During the first wave of the pandemic in Norway, many mental health facilities were forced to reduce or temporarily suspend their services [4, 5], which might have affected the access and utilization of such services.

Studies from the UK found a reduction in the number of primary care consultations for mental health problems during the first period of the pandemic [6, 7] and reduced incidences of primary care-recorded mental disorders [8, 9] among the general population. When looking at a longer period during the pandemic, some of the studies from the UK found that the number of primary care consultations for mental health problems remained lower than expected by the end of the study period compared to the pre-pandemic years [6‐8]. In Norway, the number of general practitioner (GP) consultations for mental health problems increased during the spring and early summer of 2020 [10]. However, by July and August 2020, the level of consultations decreased towards pre-pandemic levels. By September 2020, the number of consultations accelerated, and the increase lasted throughout 2020.

The previous studies mentioned above have only investigated the first few months or the first year of the pandemic and focused on service utilization in the general population. During a public health crisis like the COVID-19 pandemic, it is also important to identify and investigate vulnerable groups who may be at greater risk of experiencing worse psychological outcomes [11]. Limited access to mental health services and the negative psychosocial effects of social distancing measures are likely to disproportionately affect individuals with a history of mental health problems [12, 13]. For instance, limited access to services or fear of seeking help could increase the risk of discontinuation or abrupt termination of mental health treatment, which can worsen mental health problems [14]. Lack of social support and activities that maintain mental health could further exacerbate psychiatric distress and affect overall functioning [15]. In addition, individuals with pre-existing mental health problems are also more susceptible to stress than the general population [16], which could increase relapse rates [17, 18]. A systematic review and meta-analysis based on data-collections done in 2020 did not find any evidence of an increase in symptoms of mental disorders among individuals with a pre-existing mental disorder [19]. However, it is unknown whether these findings can be generalized to the prolonged duration of the pandemic, when the strain on people with mental health problems may have been even greater. Consequently, it is important to investigate whether this group’s mental health care use changed during the pandemic.

To our knowledge, only a few studies have investigated the effect of the pandemic on service utilization among individuals with pre-existing mental health problems. Ridout et al. [20] used electronic health records in a large community, primarily an employer-based health system in the US, and investigated psychiatric service utilization in 2020 compared with 2019. They found that the number of patients with a pre-existing psychiatric diagnosis seeking help remained stable during the first year of the pandemic. In contrast, Janoczkin et al. [21] analyzed individuals presenting to one Emergency Department in America with psychiatric complaints between January 1-July 9 in 2019 and 2020 and found an increased prevalence of patients presenting with a history of previous psychiatric care. In addition to that these two studies have solely been based on data from 2020, another important limitation is their small and limited subject samples.

There is hence an important knowledge gap in the literature, as no published studies so far have used whole-population registry data to investigate the utilization of primary care for mental health problems during a prolonged period of the pandemic among adults with a recent history of mental health problems.

In Norway, all citizens registered as living in Norway in the National Population Register have the right to a General Practitioner (GP). GPs could be seen as gatekeepers to other health services, as they often are the patient’s first contact within the health care system, and the patient may contact their GP without a referral. Based on the patients’ needs, the GP can refer the patient to services in specialist health care, such as hospitals and outpatient clinics owned or financed by the government. In Norway, the treatment of moderate to severe mental disorders is given by specialist health care, and patients need a referral for help. In Norway, primary and specialist health care are linked treatment services, and increased access to specialist health care may lead to less demand for primary care and vice versa. During the pandemic, all GP offices remained open, although with strict restrictions regarding in-person consultations. At the start of the pandemic lockdown in March 2020, mental healthcare facilities were open for emergency care only, and patients were advised to wait or have electronic consultations (phone/video). However, the mental health care services adjusted their services to accommodate the pandemic situation and the level of restrictions.

The present study aims to estimate to what extent the number of primary care consultations for common mental health problems changed during the pandemic (2020 and 2021) compared to the pre-pandemic years (2018–2019) among individuals with a pre-existing mental health problem. To investigate changes in consultations before and during the pandemic, we created two different data sets (one for each time period), containing GP consultations for mental health problems among individuals with pre-existing mental health problems. We chose to focus on consultations for depressive disorder, anxiety disorder, phobia/obsessive–compulsive disorder (OCD), post-traumatic stress disorder (PTSD), and anorexia nervosa/bulimia (eating disorders), as these disorders have been hypothesized to be especially affected by the increased stress related to the COVID-19 pandemic [22‐24]. The present study also adds to the existing literature by examining differences in service utilization patterns across disorders groups.

Our study, which includes nationwide registry data from 2017–2021, indicates that the COVID-19 pandemic has differently affected service utilization for various mental health problems. Among individuals with pre-existing mental health problems, we found a massive increase in primary care consultations for PTSD and eating disorders during the pandemic compared to the pre-pandemic years. The increase in consultations for PTSD was most apparent among the youngest age group, while the increase in consultations for eating disorders was apparent among women and all age groups. We also found that the number of consultations for anxiety disorder were higher than predicted in time periods following strict social distancing measures. Further, we found no or minor changes in the number of consultations for depression and phobia/OCD.

Our results show that the number of GP consultations for mental health problems among individuals with a history of mental health problems increased or remained stable during the pandemic years. This indicates that primary care succeeded in providing and sustaining healthcare for this vulnerable group even during a national crisis. The general finding of increased use of primary care during the pandemic years among this group is in line with the study by Janoczkin et al. [21]. The increased use of primary care among some individuals with a pre-existing mental health problem might be due to several factors. Firstly, the experience of symptom worsening and risk of relapse in this group might increase the service utilization, which have been reported in several reviews [32‐34]. However, one review did not find any symptom worsening among individuals with pre-existing mental health problems [19]. Secondly, the increased number of GP consultations among our sample could be the result of displacing nurses from low threshold services for mental health care in municipalities for corona-related tasks, as this might have reduced municipal mental health care provision. Thirdly, it could be the result of reduced access to specialist health care [4], which could worsen mental health problems and increase the need for help in primary care. Lastly, the increased number of GP consultations for mental health problems might reflect that GP offices have been open during the pandemic, with increased use of e-consultations. E-consultations likely made it easier for GPs to follow their patients more regularly, which could lead to more consultations for mental health problems [35]. The new daily routines and work-situation caused by the strict confinement measures could also influence the use of GP consultations, due to more time at home, less meetings, and more flexible work-schedules. However, the present study cannot investigate any causal relationships, and it is likely that multiple factors operate at the same time.

We found a higher number of observed GP consultations than predicted for anxiety disorder after the first and second periods with strict contact-reducing measures (except in the youngest age group). In addition, we also found increased consultations during the second period with strict contact-reducing measures among women and the two oldest age groups. Our result contradicts studies that have found a reduction in the number of primary care consultations for anxiety [6, 7]. The different result might be due to the fact that Lemanska et al. [7] and Mansfield et al. [6] investigated primary care consultations for anxiety in the general population and not specifically among individuals with a history of mental health problems. The finding of increased consultations with GPs after periods with strict restrictions might indicate that some people were reluctant to use mental health services during the strict periods and that the increase was due to a spill-over effect after the strict measures were alleviated. One explanation could be that individuals were afraid of getting infected with the virus at the GP office or stayed at home due to public health restrictions [4, 32]. Since we also found increased number of consultations during the second strict period among women and the two oldest age groups, this might indicate that social distancing measures triggered or exacerbated symptoms among some, but not all, individuals, which have been previously documented [36, 37].

Further, we found a large increase in the number of consultations for PTSD in several periods during both pandemic years. The increase was observed after the first strict period, and during and after the second strict period among both genders. In addition, there were also a period with increased number of consultations during the fall of 2021 among women and all age groups. Our results specifically indicate that the youngest age group (18–24 years) have been more affected by the pandemic, compared to the older age groups. The youngest age group had the largest and longest period where the consultations exceeded the predicted levels, which lasted from April 2020-February 2021. This is in line with a Norwegian study on the general population that found that the risk of fulfilling the symptom criteria for PTSD was associated with lower age [38]. The finding of increased consultations for PTSD adds to the literature using survey data, as some studies have found that a considerable proportion of psychiatric patients have reported severe PTSD-like symptoms during the initial stage of the pandemic [39, 40]. One possible explanation for the present finding could be that individuals with pre-existing mental health problems have a higher risk of developing PTSD than individuals without a history of mental health conditions, as were documented in a Norwegian sample [41]. Therefore, the pandemic may have increased stress levels among individuals with a pre-existing mental health problem, which in turn might have triggered or exacerbated symptoms of PTSD. However, there are debates as to whether the COVID-19 pandemic can be considered a traumatic event and fulfill DSM-5's Criterion A for receiving a PTSD diagnosis [42]. It is therefore more likely that the pandemic increased symptoms among some individuals who already struggled with PTSD before the pandemic. Another explanation could be a biological effect through the COVID-19 virus itself or via inflammatory or other immune processes [43], as increased plasma concentration of the inflammatory marker C-reactive protein (CRP) are associated with PTSD symptoms [44]. However, estimates from the Norwegian Institute of Public Health suggest that less than 10% of the Norwegian population had been infected by the coronavirus by the end of 2021 [45], compared to 40–70% of the population by the end of June 2022 [46].

Lastly, we found a massive increase in the number of observed consultations for eating disorders among women and all age groups. Except from the beginning of the pandemic in 2020, the end of 2021, and between February 2021-March 2021 among the youngest age group (18–24 years), the observed number of consultations exceeded the predicted number of consultations during both pandemic years. This is the opposite of what Mansfield et al. [6] found in their study on primary care consultations in the general population, but supports findings from specialized health care on service use [47‐49]. The divergent finding of Mansfield et al. [6] might be due to sample differences, as our finding is in line with studies from specialized health care, where individuals have an existing mental disorder [47‐49]. A systematic review reported symptom worsening among individuals with eating disorders during the first year of the pandemic [33], but the variation in reported symptom worsening is large [50]. When looking at the literature on children and adolescents, a Norwegian registry study found a major increase in the use of mental health services (both primary- and specialist care) for eating disorders during the first year of the pandemic [51]. Our study extends the research findings on children and adolescents and shows that the increased use of health care for eating disorders also is evident among adults with a history of mental health problems. Symptoms worsening and the increased number of GP consultations for eating disorders during the pandemic could be the result of isolation, loss of structure, changes to routines, and negative influence of social media [52].

In all time series plots, there was a sharp decline in the number of observed consultations during December 2021, while this was not the case for December 2020. This is a statistical artefact probably due to some consultations in December 2021 being registered in January 2022, and thus not included in the pandemic data set.

There are several strengths of this study. Firstly, the use of data from KUHR captures all patient encounters with publicly funded GPs in Norway, due to population coverage and specifically assigned GPs for all Norwegian citizens. Secondly, we had access to data from both 2020 and 2021, as a major limitation of the existing literature on the effects of the pandemic on mental health is the use of data from only 2020. Thirdly, our study design is based on an identical statistical procedure for the two data sets. Individuals with a pre-existing mental health problem were identified before the pandemic in both data sets (2017 and 2019). In addition, the number of consultations among these individuals was examined over two years, either before the pandemic (2018–2019) or during the pandemic (2020–2021). Fourthly, by measuring the use of primary care, we may have captured more individuals who have experienced mental health problems in our study population, as there is no need for a referral to get in contact with a GP or a risk of being rejected. This might give a better indication of how the pandemic has affected mental health compared to using data from specialist health care, which has capacity problems and is based on referrals.

Our study also has some limitations. Firstly, we only had access to data from 2017–2021. Thus, we only had one observation period before the outbreak of the COVID-19 virus. However, by using bootstrap we tried to compensate for the lack of access to more time periods before the pandemic when estimating the correlation coefficients used to make predictions. Secondly, we defined individuals as having a pre-existing mental health problem in the two data sets based on one inclusion year (2017 and 2019 respectively). Hence, we do not know if the patients defined as having a pre-existing mental health problem in our two data sets had frequent contact with GPs prior to the inclusion years. However, by relying on one inclusion year, we made the definition of pre-existing mental health problems identical in the two data sets. Thirdly, we assume that the pre-pandemic data set reflects a normal service use pattern and that the major difference between the two data sets is associated with the pandemic. In other words, we do not consider whether the mental health problems' prevalence or service use may have changed due to other factors during the pandemic. Fourthly, our prediction model is based on the assumption that the number of weekly GP consultations during the pandemic years are maximum the same or less than the consultations during the pre-pandemic years. In other words, we assume a stable development of GP consultations over time. However, it is likely that the trend of consultations is more nuanced and varies between years. Despite this, the predicted number of consultations for depression and phobia/OCD largely overlapped with the observed number of consultations, which indicates that our prediction model was able to make good predictions. Lastly, we rely on mental health problems diagnosed by GPs in routine care settings [53]. However, one study found that diagnoses in health registries (including KUHR) have moderate sensitivity and excellent specificity [54]. Moreover, we have no reason to believe that the diagnostic practice among GPs for mental health problems have changed during the pandemic.

This study demonstrates that the COVID-19 pandemic has affected service utilization for mental health problems in various ways. Among some individuals with a pre-existing mental health problem, the use of primary care increased during the two pandemic years. Specifically, individuals with anxiety, PTSD, and eating disorders have been most affected by the pandemic and may have experienced clinical deterioration and relapse. In comparison, the pattern of service use among individuals with depression or phobia/OCD generally remained unchanged during the pandemic. Our study has implications for mental health care during a future crisis, as we have identified groups of mental health problems that might need special attention during a global stressor such as a pandemic.