Association between socio-economic deprivation and AHRQ composite indicator during pandemic

Abstract Background The Agency for Healthcare Research and Quality (AHRQ) identifies preventable hospitalizations as proxy of potentially low-quality of care. Previous studies showed as socio-economic status was associated to poor diseases outcomes and to the performance of health services. Recently a particular attention was focused on the effect of the pandemic on this context. The aim of this research is to analyze the association between poor quality of primary care and socio-economic status before and during the pandemic. Methods A retrospective observational study was conducted in Abruzzo Region, Italy. Hospital discharge records (HDR) of two different periods were selected: from April to December for 2019 and 2020. The aggregate Prevention Quality Indicator 90 (PQI-90) has been coded according to the indications of the AHRQ. The Italian socioeconomic deprivation index (DI), divided in quintiles (from 1st less deprived to the 5th most deprived) was attributed to all patient, based on the municipality of residence. A multivariate logistic regression model was performed to evaluate the association between PQI-90 and DI. Results Totally were analyzed 253,063 HDR, of which 14,845 attributable to the PQI-90. By correcting for gender, age and number of comorbidities, the DI was not associated with the PQI-90 during 2019. During 2020 the PQI-90 was associated to 4th DI quintile (aOR 1.19;95%CI 1.09-1.30) and 5th DI quintile (aOR 1.13; 95%CI 1.03-1.23), compared to the 1st quintile. Conclusions The impact of the pandemic on primary care has been substantial. Compared to the pre-pandemic era, during the pandemic, an association between potentially poor quality of care and the most disadvantaged socio-economic areas has been shown in Abruzzo. This evidence must be an interesting starting point for health planning in order to fight against inequalities in health services access. Key messages • The reduction in the primary care quality during pandemic caused potentially preventable hospitalizations and it was associated to socioeconomic deprivation. • The analysis of hospital admission linked to context indicators such as the deprivation index, can be a useful tool for the policy maker in order to reduce healthcare inequalities.

Digital and data tools are fundamentally changing approaches to health and the design of health systems, but governance models have neither followed nor kept up with the pace of innovation. In response to this challenge, The Lancet & Financial Times Commission on Governing health futures 2030: Growing up in a digital world explores the convergence of digital health, artificial intelligence, and other frontier technologies with universal health coverage to support attaining the SDG 3. Children and young people are crucial groups requiring particular attention to ensure that no one is left behind in achieving universal health coverage and SDG 3 amidst the digital transformation in health. Today, there are 1.8 billion people between the ages of 10 and 24 -the largest youth population in history -90 percent of whom live in developing countries. This cohort represents an unprecedented powerhouse of human potential and digital engagement that could transform health to reach sustainable development goals. This presentation introduces several key findings from the Commission's report which pertain to the governance of (digital public) health futures amidst digital transformations in health. It will highlight how human-centred approaches to health are vital to navigating the digital transformations and maximising their benefits for population health and wellbeing. Further, it will provide an action plan for meaningful youth engagement in the design, development, implementation, and evaluation of digital public health policies, programmes, and services.

3.G. Pitch presentations: Data for health services research
Background: Preventing the spread of healthcare-associated infections (HAIs) in Intensive Care Units (ICUs) constitutes a priority for Public Health. In a country with decentralized healthcare system, the comparison between and within regions might represent a useful approach to identify what hospital and ICU indicators are associated with HAIs.

Methods:
Using data from the SPIN-UTI (''Sorveglianza attiva Prospettica delle Infezioni Nosocomiali nelle Unità di Terapia Intensiva'') network, the present analyses aimed to identify the main hospital and ICU indicators associated with HAI incidence at national level, and to stratify the analyses between Italian regions.

Results:
No associations between hospital/ICU characteristics and HAIs were evident at national level. However, ICUs in Southern Italy showed the highest incidence density of HAIs if compared with those in Central and Northern Italy (p < 0.001). Stratified analyses found a positive association of incidence density of HAIs and total days in ICU in Northern Italy (b = 0.3; SE = 0.1; p = 0.002); a positive associations with ICU size (b = 1.8; SE = 0.7; p = 0.020), total days in hospital (b = 0.06; SE = 0.02; p = 0.037) and total days in ICU (b = 0.5; SE = 0.1; p = 0.006) in Center Italy; a positive association with hospital size in Southern Italy (b = 20.3; SE = 9.4; p = 0.033).

Conclusions:
Although our study confirms that HAIs still represent an important issue in Italian ICUs, there is some variation between regions from Northern, Central and Southern Italy. In general, we found that HAI incidence increased with increasing number of beds in hospital and in ICU, as well as with the the increasing number of patient-days. However, further research is necessary to better understand if additional hospital and ICU characteristics could motivate the observed regional differences.

Key messages:
There is a large regional variation in the incidence of HAIs in Italian ICUs and hospitals. This difference that could be motivated by specific hospital and ICU characteristics.
Yet, those patients are still underrepresented in mental healthcare services and have more unmet medical needs. Although providers' bias has been well studied, up to date, little is still known about the factors explaining those biases.
We assessed the effect of general practitioners' (GPs') individual and organizational factors on their decisionmaking regarding diagnosis, treatment and referral recommendations for patient with MB with symptoms of major depression.

Methods:
An experimental study staged a video-vignette of a depressed patient with or without MB. GPs had to make decision about diagnosis, treatment and referral. We then assessed the influence of several factors on their decisions such as age, ethnicity, workload and patient confidence. ANOVA and MANOVA were used for analyses.

Results:
Overall, we found more unfavourable decisions in GPs diagnosis and treatment recommendations regarding the patient with a MB (F = 3.56, p < 0.001). In addition, they considered the symptoms of the patient with a MB as less severe (F = 7.68, p < 0.01) and would prescribe less often a medical treatment to these patients (F = 4.09, p < 0.05). Nevertheless, few factors explained these differences, except the age, the workload and the patient trustworthiness.

Conclusions:
This paper highlighted GPs biases based on apparent migration background of a patient with major depression that perpetuates ethnic inequalities in mental health care. Further research into the origins of discrimination in primary mental health care are needed to explain how and when those discriminations arise.

Background:
Socio-economic inequalities in common mental health disorders (CMDs) cut across each step in the cascade of care: (1) Less affluent individuals have a higher prevalence of CMDs, (2) are less likely to utilise treatment and (3) might benefit less from treatment when they do receive it. Here, we propose a new framework for distinguishing between these three types of inequalities in CMDs and test if such 'triple inequalities' exist globally and how they vary across countries.

Methods:
We use the Wellcome Global Monitor 2020 (N = 119,088 in 113 countries) to test if socio-economic factors, psychological factors (stigma and trust) and country-level factors (GDP, GINI and health expenditure) predict CMD lifetime prevalence, utilisation and perceived helpfulness of talking therapy and medication. Multi-level logistic regression models were used.

Results:
As predicted, people with higher household income are less likely to experience anxiety or depression (OR = 0.90 for each income quintile, p < 0.01), more likely to talk to a mental health professional (OR = 1.05; OR = 1.34 for higher