Sage Journals: Discover world-class research

Abstract

Objectives: In the early stages of the COVID-19 pandemic, evidence generation lagged behind public health responses. This study describes an international collaboration of frontline clinicians who used open data describing COVID-19 trends to generate “practice-based evidence”. Methods: Open data resources from nine Northern Periphery and Arctic (NPA) countries were harnessed using the open-source programming language ‘R' and our collaborations analyses and insights were published on a public-facing website. The website’s visualisations guided teleconference discussions from September 2020 to March 2021, focusing on contextualizing national responses, especially in rural regions. Results: This project facilitated shared learning from COVID-19 trends and highlighted key aspects of national responses. Notably, rural NPA regions experienced less COVID-19 cases and mortality in the first year of the pandemic. Conclusion: This international collaborative effort, driven by open data analysis, provided a platform to share real-world insights. The study offers a potential template for future pandemics and emphasises the importance of sustaining open data resources, including granular data like excess mortality, for effective pandemic learning.

Keywords

COVID-19 health informatics pandemics public reporting of healthcare data rural medicine

Introduction

The COVID-19 pandemic, responsible for 15 million deaths worldwide in its first 2 years,¹ generated an unprecedented global response. This response saw a public health threat and non-pharmaceutical interventions (NPIs), such as social distancing, simple hygiene measures and travel restrictions^2–4 take centrestage in mainstream public discourse for the first time in recent history. Additionally, use of large open data resources from entities such as the World Health Organisation (WHO), Google and Our World in Data (OWID)^5–9 has allowed the general public access to ostensibily comparative data to analyse the trends relating to this novel threat for themselves.

As time passes, the daily relevance of NPIs and COVID-19 data dashboards are receding from public consciousness as countries return to normality. However, COVID-19 continues as an active issue across health systems.^10,11

In the initial stages of this pandemic, by necessity, public health responses preceded evidence generation. As health systems faced unprecedented strain from COVID-19 spread amongst nonimmune populations, there was a recognition that evidence-based medicine was too slow to inform all decision-making. Accordingly, many clinicians called on colleagues to expand traditional evidence-based practice to rapidly include ‘practice-based evidence’,¹² whereby natural experiments largely provide the data for scientific study.

As the third noteworthy coronavirus outbreak of the past 20 years, SARS-CoV-2 will not be the last global public health challenge we will face.¹³ Thus, lessons must be learned from what our national responses have achieved, and where they have fallen short. Modern programming techniques, which can automatically download, filter and rebuild freely available open data into useful visualisations, can facilitate timely and ongoing analyses of public health responses and subsequent changes in healthcare system trends.^5–10 While such resources offer an opportunity to inform such approaches, there are significant caveats, including how differences in testing and reporting criteria make international comparisons difficult.^14–16 However, comparisons within national boundaries, where similar definitions and health policies have been applied over time, may offer additional insights. Mature reflection on the pandemic must also be mindful of “excess mortality”, which “measures the additional deaths in a given time period compared to the number usually expected”.¹⁷ As healthcare systems changed care pathways considerably, and public health messaging led to an understandable reluctance for many to attend healthcare settings, direct and indirect consequences of COVID-19 were, and likely continue to be, significant.

This article describes an open data, web-enabled methodology and presents key findings from a collaboration of frontline primary care, public health physicians and researchers from nine Northern Periphery and Arctic (NPA) programme¹⁸ countries. The Northern Periphery and Arctic 2014-2020 Programme formed a cooperation between nine programme partner countries: the European Union member States of Finland, Ireland and Sweden in cooperation with Northern Ireland, Scotland, Norway, the Faroe Islands, Iceland and Greenland. While the entire territories of the latter three countries are designated as “NPA” regions, all other more populous countries are divided into more rural “NPA” and more urbanised “non-NPA” regions. For example, the western seaboard of the Republic of Ireland is designated as an NPA region, while the midlands and east of the country are designated non-NPA. The NPA programme primarily focused on remote primary healthcare settings and thus analyses of subnational data, organised into more rural NPA and more populous non-NPA regions, were of particular interest. Using open data resources, analysed and visualised on a custom-built website, as the backdrop to discussions, our remote collaborative effort aimed to identify and collate data to assist us in generating practice-based evidence.

Methods

Setting

The project team comprised core members from Ireland, Iceland, Norway and Canada responsible for the original project proposal, with other NPA-associated researchers and clinicians (from the Faroe Islands, Finland, Scotland, Sweden and Northern Ireland) invited to participate via email.

Despite geographical differences, designated NPA regions within partner countries share many common features, such as low population density, low accessibility, low economic diversity, abundant natural resources, and high impact of climate change.¹⁸ Please see Online Appendix A for list of NPA countries and regions.

Data sources

Demographic, geographic, COVID-19 and excess mortality data were gathered from 18 online resources (see Online Appendix B for full list of resources). While national data are available from large online databases curated by Google⁶ and “Our World in Data” (OWID),⁷ more detailed regional data were only available through local governmental and other official sites not included in larger databases. These latter data sources were important to facilitate more detailed analyses not available on other websites.

Data harnessing and analysis

Open data resources were harnessed using ‘R’, the open-source programming language.¹⁹ Once collated, cleaned and organised, ‘R-Shiny’, which allows users to build interactive web applications using R code,²⁰ was used to create a public-facing website for collaboration partners and the general public to explore.²¹ Screenshots of the website and links for the current version of the website are included in Online Appendix C. Code written for this project and data used is available at a github repository.²²

To create meaningful comparisons, all cases and deaths data were standardised to a rate per 100,000 population, and granular daily data facilitated grouping by time according to user preference (by day, week, month). Excess mortality data is presented as percentage change between the national total number of deaths in each month, as compared to the aggregated baseline data for the period 2015-2019, from the open data resources detailed in Online Appendix B.

In addition, various sections of our website were created to explain key public health concepts important for communication with the general public and other clinicians, illustrated with real-world data, at national and regional level (where available). These concepts included the importance of testing strategies and subsequent impact on data trends, varying definitions of COVID-19 deaths, illustration of some of these interrelated concepts through regional maps and explanations of case fatality proportion and percentage positivity.

For regional analyses within countries with available data, the Chi-squared test was employed to examine differences in proportions of case counts vs non-infected populations and death counts vs survivors across NPA and non-NPA regions.

Project reporting

Monthly teleconferences were held with our group of clinical experts involved in the NPA programme between September 2020 and February 2021. Scheduled monthly sessions involved the appraisal of latest data on our project website, highlighting trends in testing, case and death data across the partner territories, in addition to key public health developments and interventions ongoing in each country.

Our aim was to use open data to identify trends in COVID-19-related data that could be explained by speaking with local experts. Generation of practice-based evidence thus involved:

1) building a live, public-facing dashboard from open, comparative data from partner countries.

2) using this dashboard to stimulate discussion amongst collaborators.

3) Identifying and exploring key aspects of public health interventions in each territory, with a particular focus on the rural NPA regions in partner countries, to feed back to step 1, to benefit our collaboration and publish our insights in an open fashion.

Results

Analyses

At the national level, cases and deaths data were available for all countries.

In addition, regional case data, to facilitate dividing countries into NPA and non-NPA regions, was available for all countries, and thus appraisal of regional spread of COVID-19 cases was possible. Note regional data were generally reported at “county” (or other large administrative geographic area)-level by countries examined.

However, regional COVID-19 deaths data were not published in an open data format by Finland and Norway, and therefore a regional breakdown of deaths was not possible for these countries.

Analysis 1: Case data

While there likely was large underreporting of infection counts across all territories initially as testing systems were established, over the first year of the pandemic four countries examined exceeded 3000 cases per 100,000 population (Ireland, Northern Ireland, Scotland, Sweden – see Table 1). Meanwhile, the Faroes, Finland, Iceland, Greenland and Norway confirmed COVID-19 infection rates were much lower. During the second year of the pandemic cumulative total case rates ranged from 10,000 to 60,000 cases per 100,000 (i.e. 10-60% of the population) for all countries examined. Some countries, particularly The Faroes Islands, Greenland and Iceland, with small populations (53k, 57k, 340k respectively) see very large rates due to small population size effects. In any case, considerable COVID-19 infection rates were seen in countries examined from March 2021 to February 2022. Lower COVID-19 case rates were found in NPA regions as compared to non-NPA regions in all applicable countries in both the first and second years of the pandemic.

Table 1.

COVID-19 Cases per 100,000 population March 2020 to February 2022 (data by NPA vs non-NPA regions where available).

		2020										2021		Year 1											2022		Year 2
		Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Jan	Feb	Overall	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Jan	Feb	Overall
Faroe Islands⁹	All NPA	324	35	0	0	73	357	107	54	15	205	84	8	1262	6	6	104	111	397	35	340	1855	2841	4102	26593	28034	64423
Finland²³	NPA	25	17	4	1	2	8	36	73	120	74	97	108	565*	132	59	47	32	116	261	191	246	575	1075	3627	3222	9582*
Finland²³	Non NPA	63	92	45	9	6	24	60	128	300	235	216	342	1519	626	205	161	88	293	550	273	327	734	2147	5938	3252	14593
Greenland⁹	All NPA	18	2	4	0	2	0	0	5	2	16	5	0	53	2	0	16	18	121	377	425	391	1153	2027	14206	1786	20520
Iceland⁹	All NPA	334	195	3	5	18	65	183	630	155	107	73	14	1784	46	79	31	22	386	835	298	499	1298	2703	11979	18338	36514
Ireland²⁴	NPA	44	180	48	4	8	25	107	578	228	431	2062	357	4071*	190	175	217	222	622	1086	787	1236	2447	3256	8783	2513	21532*
Ireland²⁴	Non NPA	80	460	118	13	17	80	186	503	231	439	2243	545	4916	425	321	295	205	598	1019	876	1191	2732	3981	9077	2659	23379
Northern Ireland²⁵	NPA	6	107	93	13	13	62	245	1243	690	1203	1517	404	5594*	240	152	134	260	1376	2259	1928	1931	2409	5246	6916	4019	26870	p < .02
Northern Ireland²⁵	Non NPA	3	212	67	10	8	97	371	1904	681	875	1276	459	5962	259	125	86	204	2183	2311	1778	1799	2059	5216	6653	4394	27069
Norway²⁶	NPA	67	30	5	2	5	21	45	96	160	134	146	49	759*	135	131	120	91	116	332	262	293	1014	1458	5289	8713	17955*
Norway²⁶	Non NPA	116	65	18	14	10	39	71	153	380	340	303	245	1753	704	433	300	121	118	510	707	365	1279	2950	9169	8284	24939
Scotland²⁷	NPA	39	78	16	4	4	12	74	172	114	343	625	204	1684*	85	41	39	372	640	1473	1567	1159	1695	2946	4251	3660	17927*
Scotland²⁷	Non NPA	49	188	108	12	8	41	190	696	613	692	913	421	3931	324	142	210	1009	1156	1837	2427	1448	1571	4766	5406	3560	23856
Sweden²⁸	NPA	24	113	127	215	108	32	40	199	1110	1594	847	1072	5481*	1184	1943	1029	277	89	179	224	211	218	941	6081	2627	15001*
Sweden²⁸	Non NPA	39	185	155	269	104	70	79	412	1194	1994	1093	852	6446	1218	1823	799	165	111	230	292	182	292	1418	7907	2762	17199

* denotes significance p < .01 on Chi-squared testing examining cases vs non-infected populations across NPA and non-NPA regions.

Analysis 2: mortality data

Table 2 demonstrates how higher COVID-19 mortality was experienced in Ireland, Northern Ireland, Scotland and Sweden in the first year of the pandemic, chiefly during April-May 2020 and Winter 2020/21, which understandably tallies with the corresponding case data of Table 1. While Ireland, Northern Ireland, Scotland and Sweden saw their comparatively high COVID-19 mortality rates fall in the second year of the pandemic, the other five countries saw their low initial mortality rates rise. Despite significant increases in detected cases during the second year of the pandemic across all countries, COVID-19 deaths during the second winter did not rise in proportion to these case surges.

Table 2.

COVID-19 deaths per 100000 population March 2020 to February 2022 (data by NPA vs non-NPA regions where available).

		2020										2021		Year 1											2022		Year 2
		Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Jan	Feb	Overall	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Jan	Feb	Overall
Faroe Islands⁹	All NPA	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	1.9	0.0	2.0	0.0	0.0	0.0	0.0	1.9	0.0	0.0	0.0	21.1	1.9	7.7	19.2	52.0
Finland [-]	NPA	Regional data unavailable												14.0	Regional data unavailable												29.0
Finland [-]	Non NPA	Regional data unavailable												14.0	Regional data unavailable												29.0
Greenland⁹	All NPA	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	1.8	7.0	22.8	32.0
Iceland⁹	All NPA	1.5	1.5	0.0	0.0	0.0	0.0	0.0	0.6	4.1	0.9	0.0	0.0	9.0	0.0	0.0	0.3	0.0	0.0	0.9	0.0	0.0	0.6	0.6	2.7	4.7	10.0
Ireland²⁹	NPA	1.3	10.9	3.2	0.5	0.1	0.0	0.2	1.6	3.9	2.3	32.1	20.3	77.0*	4.2	1.2	0.6	0.2	0.8	2.5	5.0	6.5	6.2	4.9	6.2	4.0	42.0
Ireland²⁹	Non NPA	3.5	31.5	10.0	1.9	0.5	0.2	1.1	3.4	4.0	5.0	28.7	17.9	108.0	6.1	2.7	1.1	0.5	0.3	1.7	3.4	4.1	5.4	5.6	6.5	4.6	42.0
Northern Ireland²⁵	NPA	2.0	14.4	7.4	1.2	0.3	0.3	0.9	7.0	15.2	18.1	30.6	10.6	108.0*	3.2	1.7	0.5	0.1	1.9	8.6	10.5	7.0	9.2	5.1	6.7	5.6	60.0
Northern Ireland²⁵	Non NPA	2.3	32.0	11.9	0.6	0.3	0.0	1.5	8.2	18.0	12.5	14.6	8.4	110.0	2.0	0.9	0.3	0.0	2.9	12.5	8.2	11.4	6.7	6.1	8.4	3.8	63.0
Norway [-]	NPA	Regional data unavailable												11.0	Regional data unavailable												18.0
Norway [-]	Non NPA	Regional data unavailable												11.0	Regional data unavailable												18.0
Scotland²⁷	NPA	2.8	14.2	3.5	0.0	0.0	0.0	0.4	4.2	3.9	1.8	19.2	9.6	60.0*	1.7	0.0	0.2	0.0	2.0	2.9	7.6	7.2	5.2	3.5	7.4	6.3	44.0*
Scotland²⁷	Non NPA	3.5	27.6	16.6	1.8	0.1	0.1	0.7	8.6	19.5	18.1	30.8	17.4	145.0	4.3	0.9	0.3	1.2	4.1	3.7	10.7	10.1	7.2	5.1	10.4	6.6	65.0
Sweden²⁸	NPA	0.2	12.3	9.4	8.0	4.3	1.1	0.1	0.6	7.6	27.8	18.0	9.2	99.0*	7.0	7.3	3.0	1.3	0.6	0.3	2.6	1.8	0.9	1.0	6.8	15.4	48.0	p < .02, NPA > non-NPA
Sweden²⁸	Non NPA	2.5	28.7	15.0	7.8	2.6	0.6	0.5	1.8	11.2	28.0	20.7	7.4	127.0	4.8	6.2	3.2	0.8	0.2	0.4	1.9	1.2	1.1	2.3	8.2	12.8	43.0

* denotes significance p < .01 on Chi-squared testing examining death counts vs survivors in populations across NPA and non-NPA regions.

Significant differences in COVID-19 death rates were found between NPA regions and non-NPA regions of three of the four of the countries (Ireland, Scotland, Sweden) in the first 12 months of the pandemic. Despite persistence of significantly higher case counts across all non-NPA regions, a similar difference in death rate across NPA versus non-NPA regions only persisted in the second year of the pandemic for Scotland.

Analysis 3: excess mortality data

Table 3 shows four countries examined (Ireland, Northern Ireland, Scotland and Sweden) experienced greater than 10 percent increases in excess mortality in the first year of the pandemic. These countries, with the exception of Sweden, also experienced greater than five percent excess death rates during the second year of the pandemic. Sweden experienced the smallest increase in excess deaths in the second year of the pandemic, at just 1.2%.

Table 3.

Excess Mortality across Northern Periphery Arctic (NPA) Countries by month (%) [Baseline period 2015-2019 inclusive].

	2020										2021		Year 1											2022		Year 2
	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Jan	Feb	Overall	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec	Jan	Feb	Overall
Faroe Islands³⁰	−23.9	29.0	−50.5	16.9	−1.9	−15.2	−10.6	−2.0	22.6	−5.2	−4.9	11.1	−4.0	−36.5	29.0	−12.1	−2.6	10.8	−6.1	12.6	1.3	35.5	52.3	−2.4	57.4	10.4
Finland³¹	−0.2	8.6	5.9	5.1	1.7	3.5	7.0	2.6	6.0	4.8	−3.0	−1.3	3.3	−4.6	−0.2	6.7	5.6	10.2	12.7	12.5	13.7	15.6	15.6	11.6	12.6	9.3
Greenland³⁰	−18.8	−33.2	−6.1	16.1	−6.1	13.6	30.3	−7.5	32.8	25.5	7.1	31.8	6.7	−18.8	8.3	−36.6	1.9	44.7	21.2	11.4	24.3	−29.4	16.9	4.9	29.4	6.1
Iceland³¹	−0.9	5.1	5.9	−17.6	−8.0	−6.0	0.8	10.7	4.8	6.6	1.8	2.4	0.6	−6.5	2.0	1.5	−3.6	11.6	16.3	−12.1	−22.0	3.2	30.9	3.3	21.4	6.6
Ireland³⁰	7.1	46.9	9.9	−0.9	1.1	5.9	4.1	11.2	2.9	−1.0	23.9	18.3	11.2	−4.8	1.2	6.9	2.5	14.7	14.2	14.6	17.5	21.5	10.1	−2.3	3.2	7.8
Northern Ireland³²	5.8	40.5	21.9	0.6	2.2	8.1	9.8	15.9	25.8	13.8	20.5	2.6	14.1	−5.8	−4.5	3.4	1.0	13.1	29.1	28.3	16.9	24.9	7.4	−4.5	−5.8	8.0
Norway³¹	0.8	2.7	5.1	1.8	−1.7	2.1	0.3	−3.1	−3.1	−5.7	0.5	0.0	0.1	−13.5	−11.2	−11.9	−8.9	−2.8	5.2	1.8	13.5	34.0	32.6	12.9	11.2	5.1
Scotland³³	12.0	71.0	25.2	1.2	0.6	2.0	1.2	9.0	16.5	8.2	16.2	15.0	14.6	−3.5	0.2	25.6	−15.0	12.6	13.8	18.9	25.9	15.7	11.4	19.5	−3.3	9.6
Sweden³¹	2.0	38.9	24.5	11.2	−0.6	−0.7	−1.5	−2.8	11.5	25.1	16.1	−4.2	10.2	−7.1	−3.9	1.2	−1.5	−1.7	−0.1	3.8	0.4	3.2	4.7	−9.0	1.6	1.2

Website creation and content

These analyses of open data provided the backdrop to a public-facing website built in R-Shiny that automatically incorporated new data as new data were published.²²

Website content, covering areas such as the importance of differing testing strategies and case/death definitions, case fatality proportion, testing positivity rates, and disease trends across and within countries examined, was generated from visualisations of open data resources generated in R and minutes from our collaborative meetings.

Analysis of website traffic revealed over 4600 site visits from 1400 users from 33 countries over the period from its launch in October 2020 to February 2022.

Website contribution to collaboration insights

A custom website afforded the collaboration room to explore and explain key public health concepts, offer timely analysis and appraisal of COVID-19 trends and linked aspects of public health responses, at national and subnational level. Key “take-home” lessons from our collaboration’s discussions were posted on the website’s home page (see Table 4).

Table 4.

Main findings from NPA collaboration.

Lesson	Explanation
Communication is key	Clear, non-political communications, including statistics, and actionable instructions, can foster population investment in basic public health measures
Build and maintain public health infrastructure	Strong testing and tracing systems, built on public trust and engagement, are necessary to identify and control spread of a pathogen like SARS-cov-2
Speed of response	Acting quickly and aggressively to contain the virus, with special attention to border control, can help prevent it from spreading widely, thereby reducing the number of cases and deaths
Protect the vulnerable	COVID-19 has been shown to disproportionately affect older adults and individuals from lower socioeconomic backgrounds, and thus these groups may require special attention
Build and maintain strong data and surveillance systems	Robust data, including excess mortality data and data availability at regional level, and surveillance systems enabled by digital infrastructure can help to track the spread of the virus and identify areas that require additional resources or attention
International cooperation and data-sharing is important	Cooperation between nations can help to coordinate efforts, share resources and information. This can help assess responses to COVID-19, while building resilient health systems and preparedness for future pandemics
Emerging opportunities	While COVID-19 has severely impacted many individuals and national economies, some opportunities have emerged, particularly those linked with digitisation, new workplace practices and re-skilling

By enhancing our understanding of local and international COVID-19 trends as described above and providing a public-facing URL to share with colleagues, our website assisted our influencing of public health strategies and information campaigns across the EU as our group engaged with other clinicians, researchers and government stakeholders at various international fora and conferences.³⁴

Using key elements and screenshots of the website, largely through social media posts, we achieved widespread outreach, with over 700,000 engagements across digital platforms.³⁵

Initially the collaboration’s critical messages centred on the need for rapid responses, open data reporting, clear communication to mitigate SARS-CoV-2 spread, while also clarifying public health concepts like test positivity rates and why COVID-19 death definitions matter. We recognized the need for leadership to make informed decisions amidst uncertain data. The decentralized decision-making approach in countries like Norway and the Faroe Islands, involving local physicians, contrasted with centralized strategies in the UK, Sweden, and Ireland. Discussions evolved to emphasize protecting vulnerable groups, successful regional public-health autonomy, and the necessity for consistent official communication. Despite the obvious challenges, we also identified opportunities in NPA regions for remote work, re-skilling, and enhanced web-based collaboration.

Discussion

Main findings

A custom-built website allowed our collaboration space and scope to contextualise observed COVID-19 trends at regional and national level, in addition to highlighting many important elements of national responses. While engagement with our site was modest, “bite-size” social media posts based on its visualisations and our findings were popular. Having greater scope to explore concepts on the custom site meant we were not solely reliant on short social media posts, which can risk loss of message control and misinformation.^36,37 In addition, the site gave our collaboration reference material and a link to send to various clinicians, researchers and policymakers interested in learning about our work.

As differing testing approaches, reporting strategies and COVID-19 deaths definitions exist across countries examined, international comparisons need to be weighed carefully.¹⁶ This is clear from examination of Table 1 and Table 2, which highlight variance in case detection and death rates. For instance, the “overall” Year one and Year 2 columns in Table 2 reveal that Scottish non-NPA regions experienced the highest death rates, while Table 1 shows that their case rates were relatively modest in comparison. Differences in testing strategies, case and death definitions, and the impact of national health responses underscore the complexity and nuance required when interpreting case and deaths data.

Nevertheless, our analyses suggest countries involved in our collaboration can be divided into two broad groups based on case and death rates during the first year of the pandemic in North-western Europe (March 2020 to February 2021). Ireland, Northern Ireland, Scotland and Sweden experienced higher cases and deaths from COVID-19 during this time period compared to Norway, Finland, Iceland, Greenland and the Faroe Islands, which largely seemed to favour more regional as opposed to centralised decision-making. These associations are borne out in excess mortality data also, with the four former countries experiencing greater than 10% excess mortality in the first year of the pandemic.

Regarding our rural analyses, NPA regions experienced significantly less COVID-19 in the first year of the pandemic, although the experience of countries in the second year of the pandemic (March 2021 to February 2022), both nationally and sub-nationally, was more uniform. Large case rates were seen in all regions due to a combination of maturation of testing systems and viral spread as societal restrictions were relaxed in the wake of large vaccination campaigns.

Comparison with existing literature

Widespread use of open data resources and dashboards for information purposes has been a feature of the COVID-19 pandemic.^5–9 It has been suggested governments and researchers use these resources to focus more on recovery, future needs and ensuring measures are targeted for at-risk groups.⁸ Involvement of frontline clinicians will assist these efforts.

Analyses of COVID-19 data across countries involved in this project mirror that of much larger studies.^38,39 Whilst not perfect, excess mortality remains the best metric available to assess previous and ongoing impacts of this pandemic.^39–42 A large-scale analysis of excess mortality in Europe reveals overall excess deaths to be 12% higher than the 2016-2019 average in 2020, while in 2021 the corresponding figure was 14%.³⁹ Comparison with our analyses suggests the countries examined in this project fared better in the second year of the pandemic compared to other European countries. This is likely explained by particularly high excess mortality in Eastern Europe in the second year of the pandemic.^40,41 Regional data analyses and lower COVID-19 impacts seen in NPA regions in this project also mirror trends observed in Italy⁴³ and America.⁴⁴ While it might be expected that older rural populations might fare worse from a disease like COVID-19, some evidence points towards better social connectedness and “community spirit” amongst rural populations, which might lead to better protection of older community members.⁴⁵ Our analyses demonstrate that even across the limited group of countries examined, some countries’ NPA areas saw a greater protective effect than others. While the association in Northern Ireland was weak, the association in Scotland was stronger, which suggests NPA designation alone does not fully explain varying rates of COVID-19 activity seen in this project.

From a practice-based evidence perspective, our collaboration has highlighted several items relevant to ongoing COVID-19 challenges and other future significant public health issues. These are broadly in keeping with other studies in this vein, with calls for greater “synergy between the medical, political and population factors”,⁴⁶ consistent, coordinated and reliable information emanating from a trusted source,⁴⁷ acknowledgement of the importance of responding quickly and building contact tracing capacity rapidly⁴⁸ and a need to improve IT and public health generally to improve pandemic preparedness.⁴⁹

Strengths and limitations

Resources generated, chiefly the public-facing website and its featured analyses and tables, have helped our collaboration’s members over the past 3 years to formulate their assessments of their own country’s response to COVID-19, at local and international meetings. It is intended that this article and the code supplied should encourage and enable other like-minded researchers and frontline clinicians to engage in similar activities in the future.

While intra-country analyses are not immune to changes in testing and reporting strategies, regional analyses should be more robust than international comparisons as each area is subject to the same protocols at any given time. Analysis of regional data suggests there were significantly less cases and deaths from COVID-19 in NPA versus non-NPA regions in the first year of the pandemic. These associations, particularly for mortality, disappeared in the second year of the pandemic.

The countries involved in this project are quite a limited group, with well-developed healthcare systems, good quality data reporting, high levels of governmental trust and high vaccination rates. Thus, our findings may not reflect experiences elsewhere. In addition, NPA regions can be large and may contain both urban as well as rural regions, which can also be said of the non-NPA regions in the same countries. Our analyses are thus of a heterogeneous group and this may be reflected in the differing strength of association between NPA designation and COVID-19 impact. Given this heterogeneity, deeper analyses including examination of rurality, age-dependency, social deprivation and social capital and how they might impact on health and illness in our communities when responding to a pandemic threat are warranted.⁵⁰ This work will require granular regional data.

Analyses were affected to a small degree by some data quality issues. Case data are not available for Svalbard and Jan Mayen (NUTS region SJ), which are part of the NPA region for Norway. Regional case and death data reporting in the very early stages were unreliable for Sweden, Scotland and Northern Ireland, although these soon tally with national data, presumably due to improved reporting capabilities. From a data mapping perspective, all but one of the NPA-designated regions map directly onto boundaries that tally with regional level COVID-19 reporting. The exception was one Scottish region- UKM62 (Inverness, Nairn, Moray, Badenoch and Strathspey), which is divided between the NHS Highlands and Grampian regions.

Implications for policy and research

This article demonstrates how open data helped deliver insights, direction and a surer footing for our collaboration in the face of the new threat of COVID-19. While open data is freely available, collating it, particularly at regional level, with oversight and insight from a group of experts distributed internationally required significant effort. Standardisation of the data being collected by countries for public health reporting for illnesses like COVID-19 and in general would reduce the analytical burden of this effort, in addition to strengthening the conclusions that could be drawn from it.

Discontinuation of open data initiatives relating to COVID-19 is a concern,⁵¹ with large open data initiatives like the Google COVID-19 Open Database (COD) and John Hopkins University ceasing gathering of new data since late 2022/early 2023.^6,9 This will hinder our ability to assess the effectiveness of COVID-19 responses and thereby impair our ability to respond appropriately to future pandemics. While excess mortality is a very important metric, we also need to be able to track ongoing vaccination efforts and COVID-19 morbidity and mortality to avoid incorrect attributions.

In addition, open data initiatives can provide insights into a country’s overall healthcare status. The absence of developing countries in the World Mortality dataset also requires attention,³⁰ especially considering the severe impact of COVID-19 seen in developing countries with available data.⁵¹ Improving data collection and reporting, which will improve our understanding of how countries are faring with respect to COVID-19 and other public health issues, is critical if we are to respond in a fairer and more equitable way to future public health challenges.^52,53

Finally, while “the creation and dissemination of easy-to-navigate platforms with evidence-based data” has been highlighted as a means to counter misinformation,⁵⁴ these same platforms may also introduce risk of misinterpretation from the general public, such as beliefs that COVID-19 was “no more dangerous than the flu”.⁵⁵ In addition to navigating these challenges, the scientific community should avoid creating additional data sources and silos during an “infodemic”.⁵⁶ Larger coordinated efforts of transnational cooperation based on the method described herein could be rapidly deployed and coordinated to deliver refined results to participants and the general public quickly, which may help avoid some of the mistakes and misinformation witnessed during the COVID-19 pandemic.^54–56

Conclusions

Our transnational collaboration, formed at the onset of a public health emergency, was enabled by teleconferencing and open data to generate timely insights pertaining to national and regional impacts of COVID-19. These insights, presented on a public-facing website, extended our reach and influence with regard to public health messaging and assessments of our respective health services.

Ongoing appraisal of national responses in an open way is scientifically sound, given the huge efforts made by the general public worldwide in response to COVID-19 in the past 4 years. Such exercises in transparency and accountability will assist in ensuring public confidence for future pandemics, which may again require large public-enacted responses.

Ensuring ongoing availability of COVID-19 and excess mortality data, ideally at regional level, in addition to other important disease and demographic trends, is critical for ongoing evaluation of national responses and excess mortality trends. Standardising data collection and reporting systems across countries and promoting information sharing and open source data harnessing tools will improve comparisons, while strengthening future pandemic preparedness.

Supplemental Material

Supplemental Material - COVID-19 open data: An ecological study and international collaboration examining pandemic trends in Northern Periphery arctic countries

Supplemental Material for COVID-19 open data: An ecological study and international collaboration examining pandemic trends in Northern Periphery arctic countries by Uet Michael E. O’Callaghan, Monica Casey, Dana Pearl, Olivia Hickey, Anette Fosse, Sigurður E. Sigurðsson, David W. Savage, Katri Vehviläinen-Julkunen, Kirsi Bykachev, Anndra Parviainen, Holly Parker, Joan Condell, Gerry Leavey, Nigel Hart, Pál Weihe, Maria S. Petersen, Liam Glynn in Health Informatics Journal.

Footnotes

Acknowledgements

Thank you to all our partner country expert participants for their input and feedback into this project and particularly our forum of clinical experts from across the NPA region. The NPA COVID-19 Rapid Response group was founded by Dr David Heaney who very sadly passed away during the pandemic before he could see this and many other related projects come to fruition. We miss our friend and colleague dearly and this work is dedicated to his memory.

Authors’ contributions

MO'C, MC, DP, OH, AF, SES, DS, KVJ, KB, AP, HP, JC, GL, NH, PW, MSP & LG all contributed to the study conception and design and participated in the collaboration’s teleconferences. All (open) data sources were appraised by DP, OH and MO'C and subsequently brought for consideration by the group. Local context was provided by MO’C, MC & LG for Ireland, AF for Norway, SES for Iceland, KVJ, KB & AP for Finland, JC, GL & NH for Northern Ireland and by PW & MSP for the Faroes. Data collection, analysis and visualisation were performed by MO’C, with feedback from the group during successive teleconferences. Code was written by MO’C. The first draft of the manuscript was written by MO’C. MO'C, MC, DP, OH, AF, SES, DS, KVJ, KB, AP, HP, JC, GL, NH, PW, MSP & LG commented on previous versions of the manuscript, in addition to all reading and approving the final manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This work was supported by the Northern Periphery Arctic (NPA) Programme COVID-19 Rapid Response Call (Project #411, Reference 304-6311-2020, August 25, 2020). The NPA Programme was an initiative funded by the European Regional Development Fund (ERDF).

Ethical statement

ORCID iDs

Michael E. OCallaghan

Kirsi Bykachev

Supplemental Material

Supplemental material for this article is available online.

References

Msemburi

Karlinsky

Knutson

, et al. The WHO estimates of excess mortality associated with the COVID-19 pandemic. Nature 2023; 613: 130–137. DOI: 10.1038/s41586-022-05522-2.

Haug

Geyrhofer

Londei

, et al. Ranking the effectiveness of worldwide COVID-19 government interventions. Nat Human Behav 2020; 4: 1303–1312. DOI: 10.1038/s41562-020-01009-0.

Machida

Nakamura

Saito

, et al. Adoption of personal protective measures by ordinary citizens during the COVID-19 outbreak in Japan. Int J Infect Dis 2020; 94: 139–144. DOI: 10.1016/j.ijid.2020.04.014.

Chen

Allot

. LitCovid: an open database of COVID-19 literature. Nucleic Acids Res 2021; 49: D1534–D1540. DOI: 10.1093/nar/gkaa952.

World Health Organization . WHO coronavirus (COVID-19) dashboard > cases [dashboard]. Available from: https://data.who.int/dashboards/covid19/cases, Accessed 2024 Jun 15.

Wahltinez

, et al. COVID-19 Open-Data: a global-scale spatially granular meta-dataset for coronavirus disease. Available from: https://goo.gle/covid-19-open-data, Accessed 2023 Mar 18.

Mathieu

Ritchie

Rodés-Guirao

, et al. Coronavirus pandemic (COVID-19). Published online at OurWorldInData.org. Available from: https://ourworldindata.org/coronavirus, Accessed 2024 Jun 15.

Organisation for Economic Co-operation and Development (OECD) . Open Data in action: initiatives during the initial stage of the COVID-19 pandemic. Available from: https://www.oecd.org/gov/digital-government/open-data-in-action-initiatives-during-the-initial-stage-of-the-covid-19-pandemic.pdf (2021), Accessed 2023 Feb 11.

Dong

Gardner

. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect Dis 2020; 20(5): 533. DOI: 10.1016/S1473-3099(20)30120-1.

10.

Mathieu

Ritchie

Rodés-Guirao

, et al. Coronavirus pandemic (COVID-19). Published online at OurWorldInData.org. Available from: https://ourworldindata.org/covid-hospitalizations, Accessed 2024 Jun 15.

11.

The Lancet Editorial . The COVID-19 pandemic in 2023: far from over. Lancet 2023; 401(10371): 79.

12.

Greenhalgh

. Will COVID-19 be evidence-based medicine’s nemesis? PLoS Med 2020; 17(6): e1003266. DOI: 10.1371/journal.pmed.1003266.

13.

Peeri

Shrestha

Rahman

, et al. The SARS, MERS and novel coronavirus (COVID-19) epidemics, the newest and biggest global health threats: what lessons have we learned? Int J Epidemiol. 2020 Jun 1; 49(3): 717–726. DOI: 10.1093/ije/dyaa033.

14.

Fitzpatrick

. The challenges of international comparisons of COVID-19. Ir J Med Sci 2021; 190: 483–484. DOI: 10.1007/s11845-020-02370-9.

15.

Schellekens

Sourrouille

. The unreal dichotomy in COVID-19 mortality between high-income and developing countries. Brookings Center for Sustainable Development, 2020. Available from: https://www.brookings.edu/blog/future-development/2020/05/05/the-unreal-dichotomy-in-covid-19-mortality-between-high-income-and-developing-countries/ (Accessed 2023 Feb 11).

16.

Karanikolos

McKee

. How comparable is COVID-19 mortality across countries? World Health Organization 2020; 26(2): 45–50, Available from: https://apps.who.int/iris/handle/10665/336295 (Accessed 2023 Apr 4).

17.

Krelle

Barclay

Tallack

. Understanding excess mortality: what is the fairest way to compare COVID-19 deaths internationally? The Health Foundation. Available from: https://www.health.org.uk/news-and-comment/charts-and-infographics/understanding-excess-mortality-the-fairest-way-to-make-international-comparisons (Accessed 2023 Mar 16).

18.

Northern Periphery and Arctic Programme Secretariat Copenhagen, Denmark . NPA 2014-2020 cooperation programme document. Available from: https://www.interreg-npa.eu/fileadmin/Programme_Documents/Approved_Cooperation_Programme_Jan2016.pdf, Accessed 2023 Feb 11.

19.

R Core Team . R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing, 2021. Available from: https://www.R-project.org/ (Accessed 2023 Mar 17).

20.

Chang

Cheng

Allaire

, et al. Shiny: web application framework for R. R package version 1.7.4.9002. Available from: https://shiny.rstudio.com/, Accessed 2023 Mar 17.

21.

O’Callaghan

. COVIDWATCH-NPA project website. Available from: https://covidwatcheu-npa.shinyapps.io/covid/, Accessed 2024 Jun 15.

22.

O’Callaghan

. COVIDWATCH-NPA project (2020-2023). GitHub repository, 2024. Available from: https://github.com/mikeocLimerick/COVIDWATCH-NPA (Accessed 2024 Jun 15).

23.

Finnish Institute for Health and Welfare . Situation update on coronavirus. Available from: https://thl.fi/en/web/infectious-diseases-and-vaccinations/what-s-new/coronavirus-covid-19-latest-updates/situation-update-on-coronavirus (2024), Accessed 2024 Jun 15.

24.

Department of Health, Government of Ireland . COVID-19 HPSC detailed statistics profile. Available from: https://data.gov.ie/dataset/covid-19-hpsc-detailed-statistics-profile (2023), Accessed 2024 Jun 15.

25.

Northern Ireland Executive . Coronavirus (COVID-19). 2022. Retrieved from: https://www.nidirect.gov.uk/campaigns/coronavirus-covid-19. Dataset discontinued May 2022 – Latest data available from project GitHub repository [22].

26.

Public COVID-19 Data for Norway (covid19data.no) . COVID-19 data. Available from: https://github.com/thohan88/covid19-nor-data, Accessed 2024 Jun 15.

27.

Scottish Government . Coronavirus (COVID-19): trends in daily data. 2023. Retrieved from: https://www.gov.scot/publications/coronavirus-covid-19-trends-in-daily-data/. Dataset discontinued Feb 2023 – Latest data available from project GitHub repository [22].

28.

Public Health Agency of Sweden . Statistics on COVID-19. Available from: https://www.folkhalsomyndigheten.se/the-public-health-agency-of-sweden/communicable-disease-control/covid-19/statistics/ (2024), Accessed 2024 Jun 15.

29.

Central Statistics Office, Ireland . COVID-19 deaths and cases statistics. Available from: https://www.cso.ie/en/releasesandpublications/ep/p-covid19/covid-19informationhub/health/covid-19deathsandcasesstatistics/ (2022), Accessed 2024 Jun 15.

30.

Karlinsky

Kobak

. Tracking excess mortality across countries during the COVID-19 pandemic with the World Mortality Dataset. Elife 2021; 10: e69336. DOI: 10.7554/eLife.69336.

31.

Eurostat . Excess mortality tables. Available from: https://ec.europa.eu/eurostat/databrowser/view/, Accessed 2024 Jun 15.

32.

Northern Ireland Statistics and Research Agency (NISRA) . Excess mortality and COVID-19 deaths in Northern Ireland. Stat Bull 2022, Available from: https://www.nisra.gov.uk/sites/nisra.gov.uk/files/publications/Excess_mortality_and_Covid-19_deaths_in_Northern_Ireland_-_August_2022.pdf (Accessed 2024 Jun 15).

33.

National Records of Scotland . Deaths involving coronavirus (COVID-19) in Scotland. Available from: https://www.nrscotland.gov.uk/covid19stats (2022), Accessed 2024 Jun 15.

34.

Heaney

. Northern Periphery and arctic programme 2014-2020 – NPA COVID-19 response (CoRe): citizen summary. Available from: https://core.interreg-npa.eu/subsites/core/CoRe_-_Citizen_Summary.pdf (2021), Accessed 2024 Jun 22.

35.

Northern Periphery and Arctic Programme 2014-2020 . COVIDWATCH-EU NPA project #411 – NPA COVID-19 response group: citizen summary. Available from: https://core.interreg-npa.eu/subsites/core/1_-_COVIDWATCHEU-NPA_-_Citizen_Summary.pdf (2021), Accessed 2024 Jun 22.

36.

European Centre for Disease Prevention and Control . Social media strategy development – a guide to using social media for public health communication. Stockholm: ECDC, 2016.

37.

Gunasekeran

Chew

Chandrasekar

, et al. The impact and applications of social media platforms for public health responses before and during the COVID-19 pandemic: systematic literature review. J Med Internet Res 2022; 24(4): e33680. DOI: 10.2196/33680.

38.

Iftekhar

Priesemann

Balling

, et al. A look into the future of the COVID-19 pandemic in Europe: an expert consultation. Lancet Reg Health Eur 2021; 8: 100185. DOI: 10.1016/j.lanepe.2021.100185.

39.

Office for National Statistics (ONS) and Government Office for Science (GOScience) . Comparing different international methods of measuring excess mortality. Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/comparingdifferentinternationalmeasuresofexcessmortality/2022-12-20 (2022), Accessed 2023 Apr 2.

40.

Eurostat . Statistics explained: excess mortality – statistics. Available from: https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Excess_mortality_-_statistics#Excess_mortality_in_the_EU_between_January_2020_and_September_2022, Accessed 2023 Apr 7.

41.

Helleringer

Queiroz

. Commentary: measuring excess mortality due to the COVID-19 pandemic: progress and persistent challenges. Int J Epidemiol 2022; 51(1): 85–87. DOI: 10.1093/ije/dyab260.

42.

Németh

Jdanov

Shkolnikov

. An open-sourced, web-based application to analyze weekly excess mortality based on the Short-term Mortality Fluctuations data series. PLoS One 2021; 16(2): e0246663. DOI: 10.1371/journal.pone.0246663.

43.

Agnoletti

Manganelli

Piras

. COVID-19 and rural landscape: the case of Italy. Landsc Urban Plann 2020; 204: 103955. DOI: 10.1016/j.landurbplan.2020.103955.

44.

Hamidi

Sabouri

Ewing

. Does density aggravate the COVID-19 pandemic? J Am Plann Assoc 2020; 86(4): 495–509.

45.

Henning-Smith

Ecklund

Kozhimannil

. Rural-urban differences in social isolation and its relationship to health. Innov Aging 2018; 2(suppl_1): 770.

46.

Lupu

Tiganasu

. COVID-19 and the efficiency of health systems in Europe. Health Econ Rev 2022; 12(1): 14. DOI: 10.1186/s13561-022-00358-y.

47.

Desborough

Dykgraaf

Phillips

, et al. Lessons for the global primary care response to COVID-19: a rapid review of evidence from past epidemics. Fam Pract 2021; 38(6): 811–825. DOI: 10.1093/fampra/cmaa142.

48.

Chung

Marlow

Tobias

, et al. Lessons from countries implementing find, test, trace, isolation and support policies in the rapid response of the COVID-19 pandemic: a systematic review. BMJ Open 2021; 11: e047832. DOI: 10.1136/bmjopen-2020-047832.

49.

DeSalvo

Hughes

Bassett

, et al. Public health COVID-19 impact assessment: lessons learned and compelling needs. NAM Perspect 2021; 2021. DOI: 10.31478/202104c.

50.

Fraser

Page-Tan

Aldrich

. Social capital's impact on COVID-19 outcomes at local levels. Sci Rep 2022; 12: 6566. DOI: 10.1038/s41598-022-10275-z.

51.

Levin

Owusu-Boaitey

Pugh

, et al. Assessing the burden of COVID-19 in developing countries: systematic review, meta-analysis and public policy implications. BMJ Glob Health. 2022 May; 7(5): e008477. DOI: 10.1136/bmjgh-2022-008477.

52.

European Observatory on Health Systems and Policies . COVID-19 health system response monitor (HSRM). Available from: https://eurohealthobservatory.who.int/monitors/hsrm/, Accessed 2023 Apr 2.

53.

The Wellcome Trust . From equality to global poverty: the COVID-19 effects on societies and economies. Available from: https://wellcome.org/news/equality-global-poverty-how-covid-19-affecting-societies-and-economies (2020), Accessed 2023 Apr 2.

54.

Borges do Nascimento

Pizarro

Almeida

, et al. Infodemics and health misinformation: a systematic review of reviews. Bull World Health Organ 2022; 100(9): 544–561. DOI: 10.2471/BLT.21.287654.

55.

Agley

Xiao

. Misinformation about COVID-19: evidence for differential latent profiles and a strong association with trust in science. BMC Publ Health 2021; 21: 89. DOI: 10.1186/s12889-020-10103-x.

56.

Shultz

Rahtz

Sirgy

. Community, economy and COVID-19: lessons from multi-country analyses of a global pandemic. Springer, 2022. DOI: 10.1007/978-3-030-98152-5.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.57 MB