Simple measurements such as blood pressure, heart rate and blood oxygenation can help GPs better assess the risk of sepsis, a life-threatening condition. However, measuring inflammation blood biomarkers does not help to further improve this assessment. This was shown in research by University Medical Center (UMC) Utrecht and Radboudumc, which was published this week in the medical journal British Journal of General Practice.
Sepsis is caused by a disrupted immune response to an infection that causes one or more organs to stop functioning properly. Usually a bacterial infection, such as pneumonia, urinary tract infection, or erysipelas, is the cause, but sepsis can also be caused by a virus (as in severe COVID-19). Common symptoms include fever, shortness of breath, and confusion. Immediate treatment of sepsis is important to prevent serious complications, such as intensive care admission and death. To do this, it is important that sepsis is recognized as early as possible and the patient is taken to the hospital emergency department. Since general practitioners (GPs) are often the first to assess patients with infections, it is important that they in particular recognize the early signs and refer patients to the hospital in a timely manner.
Worldwide, sepsis is one of the leading causes of mortality, responsible for 1 in 5 of all deaths. In Western countries, it usually concerns the frail elderly. “It is precisely in this group that the symptoms are often more difficult to distinguish from a normal flu or a bladder infection that can be treated at home,” says Feike Loots MD, the physician-researcher at UMC Utrecht who conducted the study. He adds, “Moreover, as a GP you want to keep these elderly people at home in their own environment whenever possible. We have therefore investigated how GPs can assessed the risk of developing sepsis with the options they have during a home visit.”
The study showed that the following characteristics provide the most important information for early diagnosis of sepsis: (1) age over 65 years; (2) fever; (3) hypotension; (4) accelerated heart rate; (5) low blood oxygen level; and (6) change in consciousness.
By entering the number of abnormal values in a prediction model, a score can be calculated that indicates the risk of sepsis. This new score gives a better prediction than already existing scores used in a hospital setting. According to Loots, it is important that the study was done with patients who were also evaluated by their GP at that time: “Scores for certain conditions can give different results in the hospital than when the GP sees the patient. Some of the patients seen by the GP are treated at home and the GP also sees patients earlier in the course of the disease, so symptoms in this group can differ from patients with sepsis in the hospital.”
Another important goal of the study was to determine whether certain inflammation blood biomarkers (lactate, C-reactive protein, and procalcitonin) that can be measured via a finger prick at the bedside can help diagnose sepsis. Several blood values were found to provide some indication, but when combined with the other measurements of the model, they did not significantly improve prediction. “We therefore advise GPs to refrain from using these blood tests to estimate whether a patient should be referred to the hospital. Sometimes in patients after measurement these inflammation blood biomarkers are still normal and then they deteriorate quickly, which can actually put GPs on the wrong track,” says Loots.
The study was conducted between June 2018 and March 2020 at the GP posts in Ede, Uden, Oss and Den Bosch among 357 adult patients who received a home visit. In both the patients who remained at home and those who were admitted, blood was taken for biomarker analysis. The study was conducted under the direction of Theo Verheij MD, PhD, professor of family medicine at UMC Utrecht. “We in the Netherlands are leading the way worldwide in research into sepsis outside the hospital,” says Prof. Verheij, “Thanks to the fact that we have well-organized GP posts, we were able to conduct this study. The staff at these GP posts helped enormously”. However, this research does not mean that the guidelines for general practitioners will be adjusted immediately. First, research will be conducted to determine whether measuring the score in clinical practice will actually yield health benefits. The researchers expect to be able to conduct this follow-up study at a large number of GP posts starting next year.
Loots FJ, Hopstaken R, Jenniskens K, et al. New clinical prediction model for early recognition of sepsis in adult primary care patients: a prospective diagnostic cohort study of development and external validation. Br J Gen Pract, online first, 2022.
