Article 2/4: “Decisions based on dubious data”
Hospital capacity largely depended on logistics management at national level, based on the desire to avoid saturation of the hospital system. The problem is that the data on hospital bed saturation were not rigorously reliable - and still are not. At a time when the Belgian Parliament is voting on a “pandemic law” that endorses the crisis management of the past 15 months and sets it up as a model, one may therefore legitimately question the relevance of restrictive health measures based on intensive care occupancy thresholds.
The carnage announced in Belgian hospitals never happened. The statistical projections on which crisis management was based relied on exponential models that have not been borne out on the ground at any time in the past fifteen months. Admittedly, ICUs[1] had a busy year, but can we speak of oversaturation? It is no secret to anyone that in normal times, intensive care units are frequently overloaded[2] in hospitals[3]. Where does this gap between such pessimistic forecasts and reality come from?
Inadequate information systems
Knowing in real time the number of beds available in hospitals is fundamental for efficiently managing the distribution of patients across the country. Yet, during the first wave, the existing information systems were “inadequate to characterise intensive care capacity and use. When the first wave of the pandemic appeared, neither the exact number of intensive care beds in our country nor the exact length of stay in hospital and in intensive care of COVID-19 patients was known [ …]
Moreover, the committee [Consultative Committee] based its capacity calculation on administrative data (i.e. the number of licensed beds[4]), whereas the important figure in a crisis is the number of operational beds and above all the number of operational intensive care beds. Indeed, a non-operational licensed bed can only be activated if the necessary staff, infrastructure and equipment are available. The committee did not have information on available hospital staff either[5]”, according to the Belgian Health Care Knowledge Centre (KCE). Its report also questions the lack of data on patients’ length of stay. Yet this is a decisive factor in judging the pressure on intensive care. For if ICU admissions systematically exceed discharges, the hospital gradually becomes congested.
The KCE report therefore unsurprisingly concludes that the HTSC[6] lacked crucial information to judge the pressure on intensive care in Belgium[7] during the first wave. On the basis of what data, then, did the government predict a major and imminent saturation of hospitals if health rules were not respected? Neither Sciensano nor the ICMS[8] were able to provide these data. Why was so much information missing?
Lack of coordination
The public services[9] set up a series of working groups and Taskforces in order to “coordinate” their actions. Some groups and structures already existed[10] while others were created to manage the crisis[[11]. Santhea[12] nevertheless points to the lack of coordination between these taskforces: “An unbelievable number of structures were set up without the links and hierarchies between them being very clear […] The uncoordinated intervention of the multiple actors led in particular to the provision of sometimes contradictory information and to an overly slow response to needs on the ground[13]”.
For example, the FPS Public Health website counts 1994 ICU beds in Belgium[14], from which some 10 to 15% of “non-operational” beds (absent staff, etc.) should be deducted. Yet the statistics published by Sciensano only take into account the number of licensed beds, which does not faithfully reflect intensive care capacity in Belgium. Variations exist depending on the type of beds taken into account. For some hospitals, the ICU bed capacity recorded[15] by the State was therefore “over- or underestimated” in relation to the number of beds available. The State would nevertheless have every interest in carrying out a serious inventory of the number of available and operational beds in order to dispel any suspicion, as in Germany where “hospitals allegedly manipulated intensive care occupancy rates[16]”.
Hospitals cheated
Basically, hospitals must reserve 15% of their ICU beds for covid-positive patients[17]. Depending on the ICU bed occupancy rate, the authorities decide whether more ICU beds need to be made available for covid patients. In addition, hospitals had to maintain a ratio (4 to 1) between the number of covid hospital beds and the number of covid ICU beds. They had to report daily to the federal regulator on the occupancy rate of their various departments.
However, about 1 in 12 patients who had a positive PCR was counted as covid but had come for another reason. Who can demonstrate today that the ICU units reserved for covid patients were saturated exclusively with covid patients[18]? The icing on the cake: some hospitals even cheated so as not to receive too many covid patients[19], but it is currently impossible to estimate the extent of the phenomenon.
Lack of transparency at Sciensano
Sciensano itself refused to communicate the raw hospital figures to the special covid committee[20]. For months, several scientists and professionals had been complaining that they did not have full access to the whole data set[21]. The lack of transparency of the information collected by Sciensano makes it impossible to draw complete and objective analyses of what happened on the ground.
Official fake news
If the country’s intensive care capacity was not precisely known in real time, how could Le Soir run the headline[22]: “Incendie à Anderlecht : seul un lit en soins intensifs était disponible à Bruxelles […] consequence of the health crisis” citing the Minister of the Interior as its source? The Belgian media quickly seized on this news, further heightening the climate of fear among the population. During this fire[23], three people required more intensive care: two people had burns, which therefore fell under the military hospital, and one person needed to find a place in an ICU. Yet, according to our information, there were still 3 places available that evening in a large Brussels hospital! If the government did not have these figures, how could it spread this “fake news”?
Dubious statistics
The authorities certainly lacked reliable figures during this crisis. Yet, while open data has long been demanded by many university researchers, data transparency is still not on the agenda. This seriously complicates the task of judging precisely the pressure actually exerted by covid patients on hospitals.
Despite public statistics tainted by the flaws mentioned above, let us nevertheless try to analyse whether hospitals experienced oversaturation compared with a pre-covid hospital year. (to be continued)
By Luca Billi
[1] Intensive care units, or ICU in English for Intensive Care Unit
[2] Average occupancy rate of acute beds in Belgium (2018): 82% Source OECD: https://www.oecd-ilibrary.org/social-issues-migration-health/occupancy-rate-of-curative-acute-care-beds-2000-and-2018-or-nearest-year_d842f2d4-en
[3] Every year, hospitals experience seasonal overloads
[4] There are three types of beds: licensed beds, physical beds and justified beds. Broadly speaking, physical beds are the number of beds physically available, licensed beds are those whose licence is issued by a federation (e.g. the Wallonia-Brussels Federation) and justified beds are calculated on the basis of the hospital’s activity and determine the budget allocated.
[5]https://kce.fgov.be/sites/default/files/atoms/files/KCE_335B_Capacite_hospitaliere_durant_pandemie_COVID-19_Synthese_0.pdf, page 21
[6] The Hospital & Transport Surge Capacity Committee (HTSC) is a body under the FPS Public Health that has been meeting since the start of the covid-19 crisis. It reports to the RMG (Risk Management Group). The HTSC is made up, among others, of Dr. Erika Vlieghe, members of the FPS Public Health and representatives of healthcare federations (GIBBIS, etc.). The task of this unit is to serve as a link between the authorities and Belgian hospitals in order to establish the broad guidelines for managing hospital capacity.
[7] “The HTSC committee lacked important information to fulfil its mission. For example, number of infected persons in the general population, number of infected patients in residential care settings, number of deaths, length of stay, number of (available and appropriate) healthcare professionals, number of non-COVID-19 patients requiring hospital care, evolution over time of the parameters, and other. In this way it was difficult for the HTSC committee to have a clear view on the (expected) pressure on the hospitals and to anticipate adequately in organising (extra) hospital capacity” https://kce.fgov.be/sites/default/files/atoms/files/KCE_335_Surge_capacity_during_COVID-19_Belgium_Report_0.pdf, page 25
[8] Federal IT platform enabling the exchange of operational and strategic information in emergency situations
[9] In particular the FPS Public Health, Food Chain Safety and Environment
[10] RMG, RAG, Sciensano, etc.
[11] HTSC in particular
[12] Employers’ federation of Walloon and Brussels care institutions from the public sector and the non-denominational, non-commercial private sector https://santhea.be/fr/
[13]https://www.santhea.be/CMS/content/images/202010/les-institutions-de-soins-face-la-covid-19.pdf, page 9
[14]https://organesdeconcertation.sante.belgique.be/sites/default/files/documents/20210526_hospital_phase_1a.pdf
[15]I.e. “licensed” beds https://organesdeconcertation.sante.belgique.be/sites/default/files/documents/20210526_hospital_phase_1a.pdf
[16]https://www.blick.ch/fr/news/monde/scandale-en-allemagne-des-hopitaux-auraient-manipule-le-taux-doccupation-des-soins-intensifs-id16590589.html?utm_medium=social&utm_campaign=share-button&utm_source=facebook&fbclid=IwAR2dK6fYeMikMkTnUPbKVhnvy2sr7PeX01mTlfw6H7Ems-EQqL2I_w9eqP4
[17] These different phases are summarised here: https://www.siznursing.be/wp-content/uploads/2020/10/201022_communique_usi_covid-19_RS.pdf
[18] I.e. patients ill with covid and not patients with a positive PCR without covid symptoms
[19] For various reasons: wanting to keep beds available in case the condition of in-house patients worsens, financial reasons, etc
[20]https://www.7sur7.be/belgique/sciensano-refuse-de-communiquer-les-chiffres-bruts-des-hopitaux-a-la-commission-speciale-covid~a7b00f07/
[21]https://www.rtbf.be/info/societe/detail_open-data-et-sciensano-manque-de-coherence-entre-le-discours-officiel-et-la-realite?id=10668706
[22]https://www.lesoir.be/367182/article/2021-04-19/incendie-anderlecht-seul-un-lit-en-soins-intensifs-etait-disponible-bruxelles, 19/04/2021
[23] About 20 to 30 people with lighter injuries
Photo source:
AdobeStock
