University of Groningen Rational clinical examination of the critically ill patient Hiemstra, Bart

(1)

University of Groningen

Rational clinical examination of the critically ill patient

Hiemstra, Bart

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Hiemstra, B. (2019). Rational clinical examination of the critically ill patient. Rijksuniversiteit Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

531658-L-bw-Hiemstra 531658-L-bw-Hiemstra 531658-L-bw-Hiemstra 531658-L-bw-Hiemstra Processed on: 5-6-2019 Processed on: 5-6-2019 Processed on: 5-6-2019

Processed on: 5-6-2019 PDF page: 209PDF page: 209PDF page: 209PDF page: 209 209

8

Summary, discussion and

(3)

(4)

Summary Overall aim

In this thesis the evidence of clinical examination of the cardiovascular system was evaluated in critically ill patients. An observational study of high methodologic quality was designed to educate physicians on the diagnostic and prognostic value of clinical examination.

Current evidence on clinical examination

Single clinical examination findings provided insufficient information for estimating cardiac output based upon a review of previous studies (chapter 2). Physicians may be incapable of estimating cardiac output based on their clinical examination, as we found accuracies that equalled the ‘flip of a coin’. The best diagnostic performance was found in structured clinical examination based on combined clinical signs. Physicians benefit from research that educates them on the combination of clinical signs with the best diagnostic accuracy.

A cohort to study simple clinical examination findings

We designed an observational study meeting contemporary methodologic standards: the Simple Intensive Care Studies-I (SICS-I). The protocolised and systematic data collection allowed inclusion of consecutive patients, and the incorporation of additional (temporary) measurements for new haemodynamic research questions and hypotheses. We presented the study design of the SICS-I in chapter 3. During the preparation of this article, adjustment for a type I error (i.e. false positive findings) was required because multiple hypotheses and outcomes were tested in our cohort. To prevent the increase of false positive findings, the threshold for statistical significance was adjusted and described in a detailed statistical plan (SAP). In the absence of SAP guidelines for observational studies, we wrote a manuscript on the recommended content in chapter 4. Our guidelines were based on the current recommended content of SAPs for clinical trials. Over 90% of the content applies equally to both study designs.

The dual role of clinical examination

The dual role of clinical examination was investigated in two manuscripts. In chapter 5, we studied the diagnostic value of clinical examination. Eight of the nineteen commonly used clinical examination findings were independently associated with cardiac output. However, clinical examination alone yielded insufficient information to calculate the exact cardiac output. On the contrary, five clinical signs were indicative for a low cardiac output: atrial fibrillation, systolic blood pressure below 90 mmHg, altered consciousness, a prolonged capillary refill time on the sternum and severe skin mottling. These clinical signs may facilitate physicians into making further diagnostic and interventional decisions.

In chapter 6 we studied the prognostic value of clinical examination. The ability of clinical examination to distinguish 90-day survivors from non-survivors was reasonable. Their prognostic ability was similar to established prognostic scores. This findings suggested that a patient’s prognosis could be determined without additional measures by simple protocolised clinical examination.

(5)

Interventions on haemodynamics

The focus shifted in chapter 7 towards inotropic interventions aimed to improve patient-centred outcomes. A systematic review with meta-analysis showed that the evidence of dopamine-use in critically ill adults with cardiac dysfunction is sparse, and of low quality due to high risks of systematic and random errors. This finding combined with comparable conclusions derived from the systematic reviews on levosimendan and milrinone, revealed a recurring pattern: there is scarce high-quality evidence on inotrope use, and there is no convincing evidence that these drugs are beneficial for the critically ill patient.

(6)

Discussion

The obtained knowledge in this thesis holds the basis necessary for more elaborate clinical or haemodynamic research questions in critically ill patients. The overarching goal is to develop a parsimonious clinical monitoring; to identify the set of clinical examination findings with strongest diagnostic and prognostic value. Eventually, these clinical signs could be used as triggers for interventions, after they have been studied in future randomised clinical trials. There are still a few hurdles to be surmounted before reaching this stage, which are addressed below by the PICO structure.

Population and heterogeneity

The patient population in the ICU is heterogenous as patients have various admission diagnoses and suffer from multiple comorbidities. Primary and secondary cardiac failure are common reasons for shock,1_{but the underlying pathologies are diverse. The common denominator in these} patients is that all diagnostics and interventions are targeted to resolve the same life-threatening problem: to prevent (further) damage and provide good quality of life after ICU discharge. The evidence of routinely used diagnostics and interventions should be investigated in unselected, preferably multicentre, cohorts of critically ill patients.

Intervention on simple triggers

Triggers for interventions targeting haemodynamic instability should be reliable, measurable in each patient, and should have a strong association with organ hypoperfusion. Information from clinical examination would be suitable as it is fast, readily available and inexpensive. We investigated the value of a one-time clinical examination, and concluded that this provided insufficient information for assessing haemodynamic status. The word ‘dynamic’ in haemodynamic instability already suggest a volatile, changing process. Future research should study the variation of clinical examination and associated interventions over time, and reassess its diagnostic and prognostic value; one of the aims of the SICS-II.2_{Machine learning techniques} on large datasets including both the triggers, instigated treatments and outcome are promising. These innovative machine learning tools allow patient clustering and identification of key clinical variables in each cluster.

A better outcome may be achieved if haemodynamic treatment is personalised by providing patient-specific treatment targets. Instead of treating patients based on population-based normal values, each patient should have individual, preferentially multivariable, treatment targets adapted to the clinical situation.3_{This requires haemodynamic profiling of patient clusters in} a healthy state, which could be achieved by technical innovations such as wearable or even implantable sensors.

Comparison of interventions

When considering the heterogeneity of ICU patients, there is a need for more specific interventions for organ hypoperfusion in each homogenous patient group. The common denominator of each intervention is to prevent (further) organ damage by optimising blood flow and oxygen delivery

(7)

to vital organs.4,5_{There is currently no high-quality evidence on which inotrope, if any, is preferred} in the treatment of patients with cardiac dysfunction.6–9_{These patients should be included in} future randomised clinical trials investigating the optimal inotrope.

In current circulatory shock guidelines the treatment of shock consists of noradrenaline and fluid administration. However, there is no high-quality evidence on the treatment target to be pursued. Since ‘supranormal’ treatment targets have shown no benefit,10–12_{inotropes are} probably of little value in patients with normal cardiac function. In a previous trial of septic shock patients, all participants received noradrenaline and were randomised to two different mean arterial pressure treatment targets.13_{While this trial failed to show any beneficial effect for any of} the two strategies, ongoing clinical trials that randomise patients based on different treatments targets will hopefully provide new insights into the optimal treatment strategy.14,15

Outcomes

Our understanding of the prognostic value of interventions would increase with future clinical trials that are sufficiently large and designed to have low risk of bias. Study endpoints should focus on improving patient outcomes and follow the Grading of Recommendations Assessment, Development and Evaluations (GRADE) approach.16_{Many trials are underpowered for addressing} multiple patient-centred outcomes, which is best mitigated through the conduct of a systematic review with meta-analysis. Although the numbers of published meta-analyses have increased exponentially, most are invalid and only few have sufficiently robust methodology.17_{There is rising} evidence that up to 30% of all interventions considered to be beneficial may be potentially false positive.18,19_{Ongoing research is evaluating the body of evidence of each intervention applied in} the ICU.20_{An indispensable tool for patients and caregivers would be to have a catalogue of all} ICU- related interventions which contains all available evidence.

(8)

Processed on: 5-6-2019 PDF page: 215PDF page: 215PDF page: 215PDF page: 215 215 The future of clinical examination

The clinical examination of contemporary patients with circulatory shock is still similar to the clinical picture described in 1941:

‘’(..) All the patients studied presented a similar clinical picture. They were stuporous or comatose. The rectal temperatures ranged from 36.1 to 41.3 degrees Celsius. The skin was pale and often covered with perspiration. The extremities were cold, and this finding usually preceded the fall in arterial pressure. The skin of the body was usually warm, although in terminal stages it too became cool. The radial pulse was feeble or impalpable. The pulse rate was rapid. (..)’’21

While clinical examination used to consist of a thorough clinical inspection, it has gradually expanded to a wide array of variables that are derived from interpreting (invasive) monitoring devices, laboratory values, electrocardiograms and radiography images. The amount of collected information will only increase in the future by innovations in monitoring devices: a handheld ultrasonography probe could become the stethoscope of the 21st_century,22_{a mobile computed} tomography (CT) or a mobile magnetic resonance imaging (MRI) may become standard monitoring in each ICU patient.23_{Caregivers are likely to suffer from an information overload,}24 and they require guidance to ensure optimal patient care.25,26

A possible solution is to provide physicians with more knowledge on the value of each collected variable in daily practice. Nowadays, caregivers are loaded by a great amount of information from clinical variables which they use for setting diagnoses, estimating prognoses or instigating interventions. Many variables probably have overlapping or even similar value, and might be redundant. Physicians may benefit from knowing the fewest number of clinical signs with the best diagnostic accurately, that entails all essential information, i.e. ‘a parsimonious clinical examination’.27

Another option is to introduce computerised clinical decision support into our daily practice.28 Novel supervised and unsupervised machine learning techniques may further unravel the value of each clinical variable. Innovative machine learning tools such as deep networks, relevance learning, and targeted learning could be used to cluster patients in more homogenous groups. Subsequently, specific targets and a likelihood of future adverse events could be computed for each individual patient.

A ‘haemodynamic utopia’ would be to provide caregivers with a bedside monitor that displays both individual organ perfusion and organ function (Figure 1). The first step would be to study the ‘dynamic’ part of haemodynamic monitoring. Contemporary ICU treatment consists of a continuous treatment evaluation by repeatedly conducting clinical examinations combined with physiological arterial blood gas values.29_{These bedside clinical examinations provide} proxies for organ perfusion and function: e.g. urine output is a proxy for renal perfusion and consciousness for cerebral perfusion. Current mobile (handheld) ultrasonographic devices allow real-time measurement of individual organ flow at the bedside and may serve as an extended

(9)

stethoscope.30_{Previous studies have already shown that ultrasonography-guided clinical} management leads to faster and more detailed diagnoses and change in treatment. Supplied by individual treatment targets obtained from machine learning tools, clinical management with a whole body ultrasonography approach may lead to better patient outcomes. This hypothesis should be tested in a large, high-quality randomised clinical trial focusing on patient relevant outcomes.

Figure 1. Parsimonious clinical examination of each individual organ

Prior to such a trial, each clinical examination and ultrasonographic finding should be studied in terms of feasibility, diagnostic accuracy and prognostic value in an observational cohort of consecutive patients.2_{Various challenges arise when studying the prognostic value of clinical} variables in an observational setting, as these variables are also (partially) collected by caregivers and are used to inform subsequent decisions for interventions. The predictive value of a clinical finding includes the value of this variable combined with the subsequent intervention(s) to

(10)

correct such a value: e.g. a systolic blood pressure is influenced by the noradrenaline dose. Since treatment strategies between physicians, hospitals and countries differ, this fact may explain why different studies identify different predictors of mortality. The prognostic value of clinical examination and ultrasonographic findings will become clearer when expanded to a multicentre or preferably even multinational setting.

In summary, a parsimonious set of clinical variables needs to be developed by evaluation of a comprehensive set of clinical examinations with bedside ultrasonography findings. Based on these variables, patients should be treated according to personalised targets obtained from machine learning. A multicentre randomised clinical trial embedded in a large multicentre registry is required to provide evidence base for these statements.

(11)

References

Lilly CM, Cody S, Zhao H, et al. Hospital mortality, length of stay, and preventable complications among critically ill patients before and after tele-ICU reengineering of critical care processes. JAMA. 2011;305(21):2175-2183.

Wiersema R, Castela Forte J, Kaufmann T, et al. Observational Study Protocol for Repeated Clinical Examination and Critical Care Ultrasonography Within the Simple Intensive Care Studies. J Vis Exp. 2019;(143), e58802.

Saugel B, Vincent J-L, Wagner JY. Personalized hemodynamic management. Curr Opin Crit Care. 2017;23(4):334-341.

Cecconi M, De Backer D, Antonelli M, et al. Consensus on circulatory shock and hemodynamic monitoring. Task force of the European Society of Intensive Care Medicine. Intensive Care Med. 2014;40(12):1795-1815.

Vincent JL, De Backer D. Circulatory shock. N Engl J Med. 2013;369(18):1726-1734.

Koster G, Wetterslev J, Gluud C, et al. Effects of levosimendan for low cardiac output syndrome in critically ill patients: systematic review with meta-analysis and trial sequential analysis. Intensive Care Med. 2015;41(2):203-221.

Koster G, Bekema HJ, Wetterslev J, Gluud C, Keus F, van der Horst ICC. Milrinone for cardiac dysfunction in critically ill adult patients: a systematic review of randomised clinical trials with meta-analysis and trial sequential analysis. Intensive Care Med. 2016;42(9):1322-1335.

Hiemstra B, Koster G, Wetterslev J, et al. Dopamine in critically ill patients with cardiac dysfunction: a systematic review with meta-analysis and Trial Sequential Analysis. Acta Anaesthesiol Scand. 2019;63(4):424-437.

van der Horst I, Keus F, Koster G, Hiemstra B, Wetterslev J, Gluud C. Dobutamine for cardiac dysfunction in critically ill adult patients: a systematic review with meta-analysis and trial sequential analysis. PROSPERO. 2016:CRD42016042829.

Heyland DK, Cook DJ, King D, Kernerman P, Brun-Buisson C. Maximizing oxygen delivery in critically ill patients: a methodologic appraisal of the evidence. Crit Care Med. 1996;24(3):517-524.

Gattinoni L, Brazzi L, Pelosi P, et al. A trial of goal-oriented hemodynamic therapy in critically ill patients. SvO2 Collaborative Group. N Engl J Med. 1995;333(16):1025-1032.

Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D. Elevation of systemic oxygen delivery in the treatment of critically ill patients. N Engl J Med. 1994;330(24):1717-1722.

Asfar P, Meziani F, Hamel J-FF, et al. High versus low blood-pressure target in patients with septic shock. N Engl J Med. 2014;370(17):1583-1593.

Hjortrup PB, Haase N, Bundgaard H, et al. Restricting volumes of resuscitation fluid in adults with septic shock after initial management: the CLASSIC randomised, parallel-group, multicentre feasibility trial. Intensive Care Med. 2016;42(11):1695-1705.

Pettilä V, Merz T, Wilkman E, et al. Targeted tissue perfusion versus macrocirculation-guided standard care in patients with septic shock (TARTARE-2S): study protocol and statistical analysis plan for a randomized controlled trial. Trials. 2016;17(1):384.

Guyatt GH, Oxman AD, Kunz R, et al. What is “quality of evidence” and why is it important to clinicians? BMJ. 2008;336(7651):995-998.

Koster TM, Wetterslev J, Gluud C, Keus F, van der Horst ICC. Systematic overview and critical appraisal of meta-analyses of interventions in intensive care medicine. Acta Anaesthesiol Scand. 2018;62(8):1041-1049.

Imberger G, Thorlund K, Gluud C, Wetterslev J. False-positive findings in Cochrane meta-analyses with and without application of trial sequential analysis: an empirical review. BMJ Open. 2016;6(8):e011890.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 CHAPTER 8

(12)

Processed on: 5-6-2019 PDF page: 219PDF page: 219PDF page: 219PDF page: 219 219 AlBalawi Z, McAlister FA, Thorlund K, Wong M, Wetterslev J. Random error in cardiovascular meta- analyses: How common are false positive and false negative results? Int J Cardiol. 2013;168(2):1102- 1107. Keus F, Imberger G, Wetterslev J, et al. Protocol for a meta-epidemiological study: Apparently conclusive meta-analyses may be inconclusive - The reliability of positive, neutral, and negative results in apparently conclusive critical care meta-analyses. Copenhagen Trial Unit. 2013.

Ebert R V, Stead EA. Circulatory Failure in Acute Infections. J Clin Invest. 1941;20(6):671-679.

Feilchenfeld Z, Kuper A, Whitehead C. Stethoscope of the 21st century: dominant discourses of ultrasound in medical education. Med Educ. October 2018.

O’Connell TW, Patlas MN. Mobile devices and their prospective future role in emergency radiology. Br J Radiol. 2016;89(1061):20150820.

Hall A, Walton G. Information overload within the health care system: a literature review. Health Info Libr J. 2004;21(2):102-108.

Szczerba RJ, Huesch MD. Why technology matters as much as science in improving healthcare. BMC Med Inform Decis Mak. 2012;12(1):103.

Duran-Nelson A, Gladding S, Beattie J, Nixon LJ. Should we Google it? Resource use by internal medicine residents for point-of-care clinical decision making. Acad Med. 2013;88(6):788-794.

Sackett DL. The rational clinical examination. A primer on the precision and accuracy of the clinical examination. JAMA - J Am Med Assoc. 1992;267(19):2638-2644.

Van de Velde S, Heselmans A, Delvaux N, et al. A systematic review of trials evaluating success factors of interventions with computerised clinical decision support. Implement Sci. 2018;13(1):114.

Saugel B, Wagner JY, Wendon J, Perel A. Getting the Full Diagnostic Picture in Intensive Care Medicine: A Plea for “Physiological Examination.” Ann Am Thorac Soc. 2015;12(11):1738-1739.

Narula J, Chandrashekhar Y, Braunwald E. Time to add a fifth pillar to bedside physical examination inspection, palpation, percussion, auscultation, and insonation. JAMA Cardiol. 2018;3(4):346-350.

19 20 21 22 23 24 25 26 27 28 29 30

(13)

220 220