Social unrest in HK spurs increase in public psychiatry service needs

The 2019 social unrest in Hong Kong is associated with an estimated 6-fold surge in probable depression and post-traumatic stress disorder (PTSD) symptoms among local adults, along with an estimated 12 percent increase in public psychiatry outpatient caseload, a 10-year study has shown.

Researchers from the University of Hong Kong (HKU) surveyed local adults from the population-based prospective FAMILY cohort between March 2009 and March 2014 (waves 1 and 2; baseline), in October and November 2014 during the Occupy Central movement (waves 3 and 4), in 2015–2017 after Occupy Central (waves 5–7), and during the social unrest related to the anti–extradition bill in June to November 2019 (waves 8 and 9), to examine population mental health burden, risk factors and healthcare needs associated with the increasingly violent social unrest in Hong Kong since June 2019. The baseline sample included 14,536–19,533 participants aged ≥10 years, while the subsequent surveys included 1,213–1,736 adults at each time point. [Lancet 2020, doi: 10.1016/S0140-6736(19)33160-5] 

Surge of depression & PTSD symptoms during 2019 HK unrest

During the social unrest between September and November 2019 (wave 9), 11.2 percent of adults surveyed had probable depression – a 5.8-fold increase compared with the prevalence of 1.9 percent in the two baseline surveys conducted before Occupy Central. According to the researchers, this increase in prevalence corresponds to an additional 590,000 adults with probable depression.

Compared with the prevalence of 6.5 percent in September 2017 (wave 7), the prevalence of probable depression saw a more than 70 percent relative increase in wave 9, corresponding to an additional 300,000 adults affected.

PTSD symptoms were present in 31.6 percent of adults surveyed between September and November 2019 – a 6.4-fold increase compared with the prevalence of 4.9 percent in March 2015 (wave 5). “This corresponds to an additional 1.9 million adults with PTSD symptoms,” the researchers noted.

The prevalence of suspected PTSD, defined as PTSD symptoms plus direct exposure to traumatic events related to the social unrest, was 12.8 percent in wave 9, corresponding to 810,000 adults affected between September and November 2019.

“[During the 2019 social unrest,] the combined prevalence of either suspected PTSD or probable depression was 21.8 percent, while the prevalence of suspected PTSD and depression comorbidity was 2.5 percent,” the researchers wrote.

The high prevalence of probable depression and suspected PTSD associated with the 2019 social unrest in Hong Kong is comparable to rates observed in individuals who experienced armed conflicts (22.1 percent) or large-scale disasters or terrorist attacks (10 percent), the researchers pointed out. [Annu Rev Publc Health 2014;35:169-183; Lancet 2019;394:240-248]

Substantial increase in service needs

“Together, probable depression and suspected PTSD would roughly add an extra 12 percent to the public sector [psychiatry outpatient] queue or equivalent,” the researchers estimated.

This caseload estimate was based on the finding that 45.7 percent of those with probable depression and 44.6 percent of those with suspected PTSD during the 2019 social unrest intended to seek professional help. Among these potential new patients, 63.9 percent would prefer to consult a medical professional. Assuming that 10 percent of these individuals would eventually require psychiatrist care, the additional caseload from probable depression and suspected PTSD would be equivalent to 3 percent and 9 percent of the public sector’s annual psychiatry outpatient caseload, respectively, based on an average follow-up frequency of every 16 weeks.

“Mental healthcare providers should plan for a substantial increase in service needs – tentatively 12 percent in excess of the current baseline, disregarding inpatient care and non-medical services,” the researchers wrote, adding that some patients whose condition was triggered by the social unrest might be unable to recover simply with a change in the external macroenvironment as the social unrest tapers.

“Hong Kong is under-resourced to deal with this excess mental health burden,” explained Professor Gabriel Leung of the School of Public Health, HKU, who co-led the study. “With only around half the per-capita psychiatry capacity of the UK, and pre-existing average public sector outpatient waiting times of up to 64 weeks, it is important that we enhance mental health and social care provision so that all those in need are able to access high-quality services.”

With fewer than half of those affected intending to seek professional care, the researchers also cautioned that careful monitoring is required to evaluate how much of the residual self-care burden would eventually become unmet needs.

Of note, more than one-fifth of the respondents with suspected PTSD during the 2019 social unrest cited privacy concerns as a reason for not intending to seek professional care. “This reflects mistrust of the authorities in accessing medical records for potential law enforcement purposes. Indeed, some people have avoided seeking medical treatment in Hong Kong due to concerns that doctor-patient confidentiality is compromised,” the researchers noted.

Risk factors of probable depression or suspected PTSD

In terms of risk factors, spending ≥2 hours per day on sociopolitical news via social media was strongly associated with both probable depression (adjusted odds ratio [aOR], 2.22; 95 percent confidence interval [CI], 1.06 to 4.62) and suspected PTSD (aOR, 2.82; 95 percent CI, 1.30 to 6.15).

Frequent use of Telegram was associated with both probable depression (daily use: aOR, 2.44; 95 percent CI, 1.31 to 4.56) and suspected PTSD (1–6 times of use per week: aOR, 1.60; 95 percent CI, 1.06 to 2.43) (daily use: aOR, 2.22; 95 percent CI, 1.35 to 3.65). Daily use of LIHKG was associated with suspected PTSD (aOR, 1.88; 95 percent CI, 1.11 to 3.17) only. 

While political views about the extradition bill and participation in rallies against the bill were unrelated to probable depression, those with a neutral view towards the bill were found to have a significantly lower prevalence of suspected PTSD (aOR, 0.39; 95 percent CI, 0.19 to 0.80).

Family support, in contrast, was found to mitigate against probable depression (medium support: aOR, 0.52; 95 percent CI, 0.33 to 0.80) (high support: aOR, 0.14; 95 percent CI, 0.08 to 0.26), but not suspected PTSD.

Study limitations

“Our estimates did not account for people younger than 18 years of age. Given that a substantial proportion of the protesters are believed to be teenagers, which is substantiated by the tip-of-the-iceberg arrest statistics of 15 percent belonging to that age group, the reported prevalence of probable depression and suspected PTSD would be the lower bound of the real population burden,” the researchers pointed out.

While the study’s sample is representative of the general adult population in Hong Kong, the researchers noted that members of the police were not purposively sampled. “The strength of the police force in 2018 … would translate into about eight officers who should have been included in our sample, assuming a similar response rate by occupation, which might be less likely given their current 12-hour shift duty rosters as part of the ‘force mobilization’,” they wrote. “We can, however, anticipate that their mental health burden would be at least that of the general population, thus potentially presenting another unmeasured downward bias of the reported estimates.”