Challenges in diagnosis and management of the septic elderly

"Frail and aged" are among the "Populations requiring special considerations in Intensive Care", Section 2.1.16 in the second edition of the CICM Syllabus for the Second Part Examination. And for some reason, from among the great list of Things That Are More Difficult In The Elderly, sepsis has appeared as the subject of Question 17 from the first paper if 2024. To be fair, there are challenges, and it is

probably reasonable to ask trainees to put them into some kind of order to reassure society that we are producing intensivists who can quickly organise information into a summary while under pressure.

That no specific expert knowledge was required for this SAQ is not a facetious statement, considering that even decades ago we noticed that our intensive care units were cultivating an increasingly octogenarian clientele, and ICU trainees often intuitively grasp these concepts very early in their training. But they probably can't articulate them until they reach a certain level of maturity, a constructivist might argue. As such, the exercise of rearranging this preexisting knowledge is actually rather high on Bloom's taxonomy, as technically it represents synthesis and evaluation.  Those who were able to do this spontaneously without much thinking were probably those who had developed a schema for this already in their long-term memory, i.e. they transformed the implicit informal knowledge into some kind of explicit organised structure in their own mind. 

And what of those who didn't? The highest mark for Question 17 was 6.5, which does not suggest a strong performance by the cohort. Moreover it appears that many missed the word "elderly" in the question.  The following structure is explicitly organised to prevent this from happening in the future:

  • Challenges of assessment of sepsis in the elderly
    • History
      • Communication barriers, altered level of baseline cognition or delirium
      • Atypical presentation: often the presenting symptoms are nonspecific (delirium, lethargy, incontinence, falls, dyspnoea)
      • Delayed presentation due to social isolation, or late detection by lay carers in an institution
    • Examination
      • ​​​​​​​Fever may be absent 
      • Anatomical site of sepsis may be harder to assess due to poor cooperation or distracting sources of  acute or chronic pain
      • Fluid balance is harder to assess (baseline poor skin turgor, chronic oedema, etc)
    • Investigations
      • ​​​​​​​Leukocytosis may be absent; procalcitonin may not be elevated; CRP may be chronically elevated 
    • Diagnostic reasoning
      • ​​​​​​​The presence of multiple comorbidities presents a distraction, as there may multiple alternative diagnostic possibilities other than sepsis
      • Polypharmacy interferes with vital signs (each masking tachycardia) and creates confounding organ dysfunction findings
  • Challenges of management of sepsis in the elderly
    • ​​​​​​​Resuscitation
      • ​​​​​​​Diastolic dysfunction is prevalent: response to fluid boluses may be unpredictable
      • Vasopressor tolerance may be limited by AF and peripheral perfusion
      • There is doubt as to the correct MAP target
      • Trial data to describe the over-80s is lacking
    • Specific management
      • ​​​​​​​Source control may be impossible as the patient's goals of care may not be served by aggressive interventions
      • Anbtibiotic therapy needs to be broader (greater risk of MRO and gram negative organism) and adjusted to impaired clearance
    • Supportive management
      • ​​​​​​​Increased risk of malnutrition
      • Higher requirements for post-ICU rehabilitiation
      • Greater risk of pressure injuries
      • Slower recovery
      • Longer hospital stay, and therefore higher risk of complications such as DVT, PE, MI, hospital-acq
    • End of life decisionmaking
      • ​​​​​​​High mortality and morbidity in this age group (age is independently associated with mortality)
      • Increased risk of organ dysfunction means higher level of more intrusive support may be required
      • Often irreversible pathology (eg malignancy) coexists with apparently reversible infection
      • Treating infection may offer symptom control at the end of life

Presenting this information in this schema was felt to be important because it matches other schema, i.e. the diagnostic section matches the usual pattern of assessment vocabulary answers for the CICM Second Part Exam, while the management section matches the puerile alphabetical system used by Deranged Physiology in management SAQ answers.  At the same time, it felt important to address the specific elements of the CICM examiner marking rubric, which identified the following items as essential:

  • Atypical presentation
  • Comorbidities
  • Polypharmacy
  • Age related physiological changes
  • Frailty, physiology and end-of-life decisions
  • Examples of these

Without overlapping much with the already existing chapter on the physiological changes of old age, any areas that risk duplicating whole sections from the above will be mostly offered in the form of links or references. Of the latter, surely the best would have to be Clifford et al (2016), which is literally titled "Challenges with diagnosing and managing sepsis in older adults". Ibarz et al (2024) is also excellent, and perhaps even better because it is free. 

Diagnostic challenges for sepsis in the elderly

There is diminished immune response, both cellular and humoral. There are fewer T-cells because of thymic involution, for example. As the aged person is not inclined to react to infection, and because so many of our diagnostic tools rely on the immune reaction to identify sepsis, the presence of sepsis is masked. For example, the elderly may present an unreliable narrative for their history, which will not be classical for sepsis:

  • Presinting symptoms will be nonspecific, for example consisting of things like confusion, lethargy, tachypnea, loss of appetite, dehydration, weakness, dizziness, falls, and incontinence.
  • The presence of an altered mental state does not suggest a CNS infection (whereas it would in the younger person); for example where urosepsis presents with confusion rather than dysuria.
  • The changed mental state makes it harder for the patient to relate accurate symptoms (i.e. you cannot effectively present your urosepsis to the ED physician if you are convinced that you are at the post office and it is the summer of 1967).
  • Social isolation or institutionalisation might lead to a delayed presentation, as either there is nobody there to detect a change in their health, or the people involved are rotating shift workers who are unable to identify it.

The background will create interference:

  • Comorbidities increase the susceptibility to sepsis (eg. diabetes, immobility, chronic vascular insufficiency, etc etc) but at the same time they decrease the likelihood that sepsis will be identified by offering plausible alternative explanations for the presentation, or by presenting the medical team with a whole range of distracting abnormalities to focus on. A classical example is HHS, which may have been precipitated by sepsis, but which occupies everyone's attention by waving an exciting high BSL in front of them.

Examination findings will be confused by comorbidities:

  • For example, fluid status assessment will be made more difficult by the chronically poor skin turgor,  presence of coexisting cardiac failure or dependent oedema due to immobility
  • Because of their lazy immune systems, the elderly may not mount a febrile response, and may actually be hypothermic with sepsis
  • Anatomic sources of sepsis may be difficult to identify because of multiple simultaneously painful body regions (eg. it is not discitis, but it looks like discitis because the patient is complaining of back pain which is in fact mainly due to the horrible hospital mattress)
  • Polypharmacy interferes with vital signs (eg. where tachycardia is blocked by chronic betablocker use, or fever is concealed by chronic steroids). 

Investigations will contain distracting findings:

  • A chronic low-grade proinflammatory state often exists in old age. Apparenly somebody was very proud of themselves for coining the term "inflammaging" to describe this. The upshot is that inflammatory markers like CRP may be chronically elevated.
  • Biochemical abnormalities, eg. deranged LFTs or impaired renal function, may either be the consequence of comorbid conditions, or may be incorrectly ascribed to them.
  • Scoring systems and diagnostic screening tools may not be validated for the population, eg. octogenarians and nonagenarians often are not represented in very large numbers in clinical trials and we do not have great quality data to base decisions on. 
  • Polypharmacy interferes with laboratory diagnosis of sepsis,  eg. where it produces confounding organ dysfunction (like LFT derangement due to chronic PPI use), or where procalcitonin is depressed, as it was for  Demir & Yılmaz (2020). 

Management challenges for sepsis in the elderly

"Everything is more difficult or slower" is a reasonable way to describe the management challenges in the septic elderly, and the rapidly ageing author is reluctant to admit that it applies to many other aspects of his advancing years. There's lots of different ways to organise these data, and the "ABCDE" approach is not the only possible one. Fortunately, much of this is already present in the chapter on the age-related changes in physiology because the challenges in sepsis are common to challenges in managing just about anything else in the elderly; but unfortunately, all of that material is very generic, and would probably not score many marks if it were reproduced in the answer to Question 17 from the first paper of 2024. These sections are listed here as links anyway, to satisfy the author's obsessive need for completeness:

That latter point made its way high into the marking rubric (in fact, where the basic standard for the answer was described) which suggests that even generic considerations would have merited an entry. One can assume that it would have been better for them to be tailored and shaped to be more about sepsis. To give these a more "septic" flavour, the following were compiled from the linked references and rearranged into a framework which should be familiar to the CICM exam candidate:

  • Resuscitation:

    • Fluid boluses need to be more careful, considering the prevalence of diastolic dysfunction. Interestingly, though this point is made in multiple review articles, they all seem to reference a completely non-sepsis-related paper on the response to exercise, which suggests that we probably just don't have the data from a septic elderly cohort. One should assume that one has no idea what to expect from a stressed elderly ventricle, and err on the side of caution (i.e. more smaller boluses with frequent reassessment). 

    • It is unknown whether blood pressure targets should be altered, but many believe that the chronically hypertensive elderly should have a higher blood pressure goal than the average adulty, and that a MAP of 65 would not be sufficient. Endo et al (2022) proposed a trial of MAP 85 for these patients, for about 72 hours. Whether this helps remains to be seen (as it hasn't before, except maybe in cardiogenic shock)

    • The added benefits of the MAP are potentially counterbalanced by the added arrhythmogenicity of the vasopressors, and the septic elderly will fibrillate readily with minimal encouragement. 

    • In general vasopressor use is limited by the capacity of the elderly cardiovascular system to sustain the challenges of increased afterload, strongly positive fluid balance, and peripheral vasoconstriction. 

  • Specific management:

    • Antibiotic management is affected by pharmacokinetic differences of old age, meaning mostly impaired clearance and the need to carefully dose adjust the antibiotics
    • Choice of antibiotics is also affected, as the elderly are weirdly prone to gram negative infections (Martin et al, 2006), which calls for broader cover
    • Source control, particularly highly invasive options, are going to be largely affected by the decisions around the limits of therapy, and this will colour the rest of the management process
    • Polypharmacy affects the management options by interfering with diagnosis (eg. where tachycardia is blocked by chronic betablocker use, or fever is concealed by chronic steroids), but also with management, where it contributes to delirium or results in unexpected drug interactions
  • Supportive management:

    • The elderly patients are more likely to be malnourished and dehydrated at presentation

    • Both the risk of DVT, and the risk from DVT, are increased

    • The soft tissue loss, poor perfusion, immobility and the inability to self-advicate makes these patients vulnerable to pressure areas

    • Secondary infections due to invasive devices ae 

    • Potentially prolonged post-ICU rehabilitation needs to be factored into the patient/family expectations

    • The loss of independence may be the outcome from a prolonged hospital admission

Which brings us neatly to:

End of life decisions in the septic elderly

First, some widely believed facts:

On one hand, aggressive management in some of the frail elderly patients is likely to cause harm and unlikely to cause benefit in terms of long term outcomes, especially where:

  • There is no plan for definitive curative therapy, eg. where there is no surgical option for source control because of poor functional status
  • The level of intervention required is highly intrisuve (eg. mechanical ventilation, dialysis, etc)
  • The patient's lifespan or quality of life is already limited by something that is unrelated to the sepsis, which means that fixing the sepsis will not improve this beyond an already poor baseline (eg. malignancy, dementia, etc 

In counterpoint:

  • End of life management does not necessarily mean "no management", and antibiotics are a completely fair intervention because they may bring symptom relief or improve the level of consciousness (Fairweather et al,  2024)
  • Sepsis is a reversible condition, and one may argue that even among the frail elderly there are edge cases where a trial of aggressive management is called for, particularly where such management is sufficiently nonintrusive that it does not pose a significant burden (for example, peripheral vasopressors, noninvasive blood pressure monitoring, etc). 

So what do you, practically speaking, do? The readers are invited to review the end parts of review articles from authority figures, such as Ibarz et al (2024) and Clifford et al (2016), to observe just how generic and noncommittal the discussion of end of life decisions is. Some of these final paragraphs may as we  have been written by ChatGPT. For example, to say that "a framework encompassing physical and cognitive status, quality of life, survival likelihood, functional performance, preferences, and treatment burden guides decisions for intensive care in older patients" is not meaningful as guidance, and one may have just as easily said "we take all the things into account". Of course we do, it's literally the most important decision about that person's life. In the broadest terms, for sepsis as well as for other conditions where there is a clearly defined and limited range of support strategies that are known to enhance survival (source control, antibiotics, early resuscitation), one can make an argument in favour of a limited trial of treatment to prevent the spiral of complications that fulfills a pessimistic prophecy, at least wherever the trial does not involve torturing the patient with unacceptably painful interventions which they are unlikely to recover from (laparotomy, etc). 

References

Clifford, Kalin M., et al. "Challenges with diagnosing and managing sepsis in older adults." Expert review of anti-infective therapy 14.2 (2016): 231-241.

Ibarz, Mercedes, et al. "The critically ill older patient with sepsis: a narrative review." Annals of Intensive Care 14.1 (2024): 6.

Martin, Greg S., David M. Mannino, and Marc Moss. "The effect of age on the development and outcome of adult sepsis." Critical care medicine 34.1 (2006): 15-21.

Bagshaw, Sean M., et al. "Very old patients admitted to intensive care in Australia and New Zealand: a multi-centre cohort analysis." Critical Care 13 (2009): 1-14.

Liang, Stephen Y. "Sepsis and other infectious disease emergencies in the elderly." Emergency Medicine Clinics 34.3 (2016): 501-522.

Aspinall, Richard, and Deborah Andrew. "Thymic involution in aging." Journal of clinical immunology 20 (2000): 250-256.

Franceschi, C., et al. "An evolutionary perspective on immunosenescence., 2000, 908." DOI: https://doi. org/10.1111/j: 1749-6632.

Martin-Loeches, Ignacio, et al. "Risk factors for mortality in elderly and very elderly critically ill patients with sepsis: a prospective, observational, multicenter cohort study." Annals of intensive care 9 (2019): 1-9.

Martin, Greg S., et al. "The epidemiology of sepsis in the United States from 1979 through 2000." New England Journal of Medicine 348.16 (2003): 1546-1554.

Martin, Greg S., David M. Mannino, and Marc Moss. "The effect of age on the development and outcome of adult sepsis." Critical care medicine 34.1 (2006): 15-21.

Li, Ling, et al. "Epidemiology of sepsis in Australian public hospitals: a mixed methods, national longitudinal study (2013-2018)." (2020).

Blot, Stijn, et al. "Epidemiology and outcome of nosocomial bloodstream infection in elderly critically ill patients: a comparison between middle-aged, old, and very old patients." Critical care medicine 37.5 (2009): 1634-1641.

Fairweather, Jack, et al. "Antimicrobial use at the end of life: a scoping review." BMJ Supportive & Palliative Care 14.e2 (2024): e1873-e1881.

Haas, Lenneke EM, et al. "Frailty is associated with long-term outcome in patients with sepsis who are over 80 years old: results from an observational study in 241 European ICUs." Age and ageing 50.5 (2021): 1719-1727.

Nasa, Prashant, Deven Juneja, and Omender Singh. "Severe sepsis and septic shock in the elderly: an overview." World journal of critical care medicine 1.1 (2012): 23.

Endo, Akira, et al. "Optimal target blood pressure in elderly with septic shock (OPTPRESS) trial: study protocol for a randomized controlled trial." Trials 23.1 (2022): 799.

Parlow, Simon, et al. "The association between mean arterial pressure and outcomes in patients with cardiogenic shock: insights from the DOREMI trial." European Heart Journal Acute Cardiovascular Care 10.7 (2021): 712-720.

Dünser, Martin W., et al. "Association of arterial blood pressure and vasopressor load with septic shock mortality: a post hoc analysis of a multicenter trial." Critical care 13 (2009): 1-7.

Honorato, Michele Ouriques, et al. "Atrial fibrillation and sepsis in elderly patients and their associaton with in-hospital mortality." Arquivos Brasileiros de Cardiologia 120 (2023): e20220295.

Martin, Greg S., David M. Mannino, and Marc Moss. "The effect of age on the development and outcome of adult sepsis." Critical care medicine 34.1 (2006): 15-21.

Demir, İsmail, and İsmail Yılmaz. "The effect of polypharmacy on procalcitonin levels in the intensive care admission of geriatric patients with sepsis." Konuralp Medical Journal 12.2 (2020): 216-222.

MUÑOZ-LOMBO, JENNY PATRICIA, et al. "Multimorbidity and geriatric syndromes. Their effect on mortality in older adults with sepsis." Acta Medica Colombiana 47.1 (2022): 7-14.

Demir, İsmail, and İsmail Yılmaz. "The effect of polypharmacy on procalcitonin levels in the intensive care admission of geriatric patients with sepsis." Konuralp Medical Journal 12.2 (2020): 216-222.