Summary and conclusions

The overall DUP decreased in the early compared with the pre-pandemic epoch. In the late-pandemic epoch, DUP-Demand persisted at this lower level, whereas DUP-Total increased. To the best of our knowledge, this is the first report to show a reduction in access delays for psychotic illness during the pandemic. We speculate that this anomalous DUP reduction was caused by earlier awareness by cohabitants of symptoms in the affected individual with whom they shared stringent restrictions against movement outside the home. This may have led to quicker help-seeking. Our prior analysis demonstrated that the single largest contributor to DUP in our region was the delay between psychosis onset and first-initiation of help-seeking, and that family members were key facilitators.^{Reference Mathis, Ferrara, Burke, Hyun, Li and Zhou10} Increasing family availability and lack of non-clinical options for help-seeking, such as school counselling resources, may have led to earlier clinical contact, thereby reducing delay. Additionally, reduced patient flow to normally crowded emergency services,^{Reference Gutovitz, Pangia, Finer, Rymer and Johnson4} with a proportional increase in presentations of psychotic disorder,^{Reference Mukadam, Sommerlad, Wright, Smith, Szczap and Solomou5} may have allowed greater attention by emergency staff to the identification and management of psychotic symptoms. Both could have led to reductions in DUP-Demand. Since most patients in the catchment received their first antipsychotic via emergency and in-patient services,^{Reference Mathis, Ferrara, Burke, Hyun, Li and Zhou10} this reduction in DUP-Demand is unlikely to be explained by the increasing availability of telehealth services.

The reduction in DUP-Total during the most stringent restrictions and subsequent increase when these were loosened support a causal role for these restrictions in reducing delays to care. However, DUP-Demand remained low across the two pandemic epochs. This may be related to persistent family availability (e.g. work from home arrangements) to homebound patients, whereas decays in DUP-Supply may reflect return to pre-pandemic clinical workflows.

Some limitations should be noted. The small sample size in the early pandemic group (n = 14) increases the likelihood of chance findings. Nevertheless, this preliminary data from an established early detection service may be of interest to other first-episode services, which may observe similar patterns. Also, this data represents changes in DUP across one primarily urban and suburban region of the USA, which may not generalise to rural areas or areas with less malleable or structural sources of delay. Although families were dominant participants in help-seeking events during the trial and in the year after, we did not have complete data for this during the entire early and late pandemic phases. Finally, the early detection campaign may have had residual effects on pathways to care, and reduced scope for further DUP reduction in this population, although this would likely have reduced the effects observed in this analysis.

These findings shed light on patterns of presentation to a specialised psychosis service during a global health crisis. Periods of external threat have previously been associated with increased social cohesion;^{Reference Salib11} our results could support Durkheim's theory that major events create group integration,^{Reference Durkheim12} which may promote early detection. It highlights the potentially positive impact of greater attention by household members on reducing delays in care for new-onset psychotic disorders. The magnitude of the reduction in DUP in response to an environmental stimulus is broadly encouraging for early detection efforts that seek to engage observers in facilitating access to care.