Aim: to evaluate if the Choosing Wisely campaign can be introduced without harming critically ill patients.

Background: Each year, around 45,000 patients are admitted to Sweden's 81 ICUs (intensive care units), costing 6-7 billion SEK (6% of total healthcare expenditure). The high costs of health care are partly attributable to overuse of diagnostic tests. Up to 30% of these tests lead to treatments that provide no benefit for patients, some of them may even be harmful(1-3). To improve quality of care while combating this problem of cost, the American Board of Internal Medicine Foundation developed the Choosing Wisely Campaign, tasking professional societies to develop lists of top five medical services patients and physicians should question. Patients in the ICU, can especially benefit from this paradigm shift. They are subjected to multiple testing on a daily - sometimes hourly - basis, therefore this field of medicine responded to the campaign quickly. Two sets of top-five-lists of tests and interventions that can be questioned have been published(4 5).

In multiple countries, including Sweden, the Choosing Wisely campaign is gathering traction among critical care physicians and is on the verge of being introduced. It is both common-sensical and will save money, so health care leaders are positive. However, the investigators have an opportunity, and a duty, to assess the evidence; can the Choosing Wisely campaign be introduced without harming ICU patients?

The present study thus aims to evaluate the introduction of the Choosing Wisely campaign in the context of the ICU. Is this change of care strategy associated with changes in 30-day-mortality? Secondary outcomes include ICU length of stay, use of non-invasive- or invasive mechanical ventilation and continuous renal replacement therapy. This is a registry-based cluster randomized controlled study (R-RCT), targeting Swedish ICU's across multiple regions. Primary and secondary outcomes will be retrieved from the Swedish Intensive care Register (SIR), making ICU participation easier and less costly.

Purpose and aims Aim: to evaluate if the Choosing Wisely campaign can be introduced without harming critically ill patients.

Background: Each year, around 45,000 patients are admitted to Sweden's 81 ICUs (intensive care units), costing 6-7 billion SEK (6% of total healthcare expenditure). The high costs of health care are partly attributable to overuse of diagnostic tests. Up to 30% of these tests lead to treatments that provide no benefit for patients, some of them may even be harmful. To improve quality of care while combating this problem of cost, the American Board of Internal Medicine Foundation developed the Choosing Wisely Campaign, tasking professional societies to develop lists of top five medical services patients and physicians should question. Patients in the ICU, can especially benefit from this paradigm shift. They are subjected to multiple testing on a daily - sometimes hourly - basis, therefore this field of medicine responded to the campaign quickly. Two sets of top-five-lists of tests and interventions that can be questioned have been published.

In multiple countries, including Sweden, the Choosing Wisely campaign is gathering traction among critical care physicians and is on the verge of being introduced. It is both common-sensical and will save money, so health care leaders are positive. However, the investigators have an opportunity, and a duty, to assess the evidence; can the Choosing Wisely campaign be introduced without harming ICU patients?

The present study thus aims to evaluate the introduction of the Choosing Wisely campaign in the context of the ICU. Is this change of care strategy associated with changes in 30-day-mortality? Secondary outcomes include ICU length of stay, use of non-invasive- or invasive mechanical ventilation and continuous renal replacement therapy as well as unplanned readmissions. This is a registry-based cluster randomized controlled study (R-RCT), targeting Swedish ICU's across multiple regions. Primary and secondary outcomes will be retrieved from the Swedish Intensive care Register (SIR), making ICU participation easier and less costly.

Survey of the field Investigations on mortality, ICU length of stay (LOS) or effect on organ support among critically ill patients after the introduction of the Choosing Wisely protocol are lacking. Evidence suggests that routine laboratory testing is associated with preventable harms, such as hospital-acquired anemia and further unnecessary testing.

Can ICUs reduce unnecessary diagnostics? A Canadian single center before- and after study aimed at reducing (unnecessary diagnostic) phlebotomy in the ICU. They showed a decrease of 6.6mL/patient-day, blood tube consumption decreased, amounting to 13276 tubes saved over 11 months. Red blood cell transfusions decreased significantly.

Study design Research questions Can the Choosing Wisely protocol, specifically minimizing the number of arterial blood gases, -venous tests and -chest x-rays, be introduced safely in Swedish ICU's?

P: Adult patients treated in Swedish intensive care units.

I: The Choosing Wisely (CW) protocol, specifically minimizing the number of arterial blood gases, number of venous tests and number of chest x-rays

C: ICU care as of today, without the introduction of the CW protocol. Each ICU is its own and others' control.

O: Primary outcome: 30-day-mortality. Secondary outcomes: ICU length of stay, time in invasive/ non-invasive mechanical ventilation, time on continuous renal replacement therapy and unplanned readmissions within 72 hours of ICU discharge.

Variables and measures There are two phases in this block randomized stepped-wedge ICU study, the control phase and the intervention phase. All participating ICU's start with a control period. During this phase, clinical monitors assess the number of arterial blood gases, -venous tests and number of chest x-rays performed at the ICU. This monitoring will assess all patients treated during a week and report tests per patient per day. As ICU's are randomized to the intervention phase the same clinical monitoring will be performed. Importantly, after six- and 12 months (irrespective of randomization allocation, if the ICU is still in the control- or intervention phase), the clinical monitor will re-assess the testing/diagnostic metrics.

Outcomes from this R-RCT will be extracted using the Swedish Intensive care Register (SIR). These include the primary outcome, 30-day-mortality and secondary outcomes: ICU length of stay, time in invasive/non-invasive mechanical ventilation, time on continuous renal replacement therapy.

Inclusion Criteria:

• Patients 18 years or older, treated in Swedish ICU's

Exclusion Criteria:

• Patients under the age of 18 years

This study uses data on a block level from participating ICUs; therefore it is not feasible to share IPD, but we may opt to share other information