Checklists to guide the supportive and critical care of tuberculous meningitis

The assessment and management of tuberculous meningitis (TBM) is often complex, yet no standardised approach exists, and evidence for the clinical care of patients, including those with critical illness, is limited. The roles of proformas and checklists are increasing in medicine; proformas provide a framework for a thorough approach to patient care, whereas checklists offer a priority-based approach that may be applied to deteriorating patients in time-critical situations. We aimed to develop a comprehensive assessment proforma and an accompanying ‘priorities’ checklist for patients with TBM, with the overriding goal being to improve patient outcomes. The proforma outlines what should be asked, checked, or tested at initial evaluation and daily inpatient review to assist supportive clinical care for patients, with an adapted list for patients in critical care. It is accompanied by a supporting document describing why these points are relevant to TBM. Our priorities checklist offers a useful and easy reminder of important issues to review during a time-critical period of acute patient deterioration. The benefit of these documents to patient outcomes would require investigation; however, we hope they will promote standardisation of patient assessment and care, particularly of critically unwell individuals, in whom morbidity and mortality remains unacceptably high.


Introduction
Tuberculous meningitis (TBM) is the most severe form of tuberculosis (TB) and causes critical illness from direct neurological injury or complications from the infection, treatment, or prolonged hospitalisation.
Due to serious complications, the management of TBM is complex, and includes supportive medical and neurosurgical measures. Currently there are no guidelines or standardised approach for the assessment and management of TBM. This is due to limited literature to develop an evidence-based approach, and because each TBM patient is unique. However, common themes exist, and a comprehensive proforma to guide patient assessment could be the first step towards standardised management.
Checklists can be powerful tools to focus attention 1 and their use in the medical field is growing [2][3][4] . We aimed to develop a comprehensive proforma for the assessment and management of TBM as well as a priorities checklist for the decompensating patient. The document cannot account for every scenario, but is designed to identify priorities; i.e. potentially reversible factors that contribute to morbidity and mortality. Local modifications to increase uptake and tailor use to suit local needs are encouraged.
Accompanying our proforma and checklist is the rationale for why these assessments may be important. Importantly, this article is not a guideline and does not make recommendations for care. It is not intended to replace a comprehensive ward round, nor to increase the clinical workload. Instead, it should provide a framework to highlight vital components during different stages of TBM care, with many complications overlapping throughout illness. We acknowledge that investigations and procedures will not be available at all centres.

Comprehensive proforma
The comprehensive proforma is split into initial evaluation (Table 1), daily inpatient review (Table 2), and critical care in the intensive care unit (ICU) ( Table 3). However, as elements can occur at any time, the rationale is grouped by themes within sections titled "General supportive and critical care" and "Neurocritical care".

General supportive and critical care
History of present illness Obtaining a thorough history of the patient's signs and symptoms is paramount (Table 1 and Box 1).

Box 1. Key predictors of poor outcome in tuberculous meningitis
Increased Medical Research Council TBM disease severity [5][6][7][8] Reduced consciousness 6,9 Hydrocephalus and raised ICP 8,10,11 Cerebral infarction 6,12 Seizures 12,13 HIV co-infection 9,14 Multidrug resistant, or isoniazid mono-resistant, disease 15 Lower body weight 5 Younger and older age 5,16 General clinical examination and monitoring General assessment and non-invasive monitoring of vital signs may be the only tools available to guide patient management in many centres. These provide valuable information in all care settings, and are mentioned in Table 1-Table 4.

Respiratory monitoring.
A change in neurological status may cause hypoxia due to airway compromise, and pulse oximetry may raise the alarm. Chest X-rays can help diagnose pulmonary TB,

Amendments from Version 1
Under respiratory monitoring we have added here that pneumothorax is rare in PJP (with a reference).
In Table 4 we have added 'preferably with contrast' in reference to brain imaging.
In Table 4 we have specifically mentioned IRIS and cerebral sinus thrombosis. In the supporting document we have added the sentence 'Cerebral venous thrombosis is an unusual cause of acute neurological deterioration in TBM, but has been described' with references.
Regarding risks of prolonged hospital admission, we have added 'Deep vein thrombosis is a risk of prolonged critical care admission' Any further responses from the reviewers can be found at the end of the article '✓' can be selected when a proforma question has been answered, or a proforma point has been reviewed or tested. Does the patient need repeat neuroimaging?

Laboratory tests (blood)
Repeat complete full blood count and inflammatory markers (i.e., if concern for other infection) '✓' can be selected when a proforma question has been answered, or a proforma point has been reviewed or tested.

General clinical examination
Weight and nutritional status (i.e., use of oral feeds, intravenous fluids, etc.) Monitor for vomiting/inability to take drugs orally

Monitor for gastrointestinal bleeding
Vital signs (i.e., oxygen saturation, heart rate, blood pressure, temperature) Hydration status (i.e., fluid input and output, clinical signs of dehydration) Monitor skin for pressure damage

Medication evaluation
Have any doses of anti-TB chemotherapy been missed?
Monitor for side effects from anti-TB chemotherapy Check drug susceptibility testing results. Are changes to anti-TB chemotherapy required?
Monitor recent liver function and renal function panels for medication toxicity

Repeat liver function and renal function panels if toxicity concerns remain
Check corticosteroid dose Schedule corticosteroid taper (i.e., when to reduce the dose)

Vascular access
Is central venous access still needed?
Is central venous access functioning properly?
Are there signs/symptoms of central line-associated blood stream infection?
Is invasive blood pressure monitoring (arterial line) still needed?

Urinary catheter
Is the urinary catheter still needed?
Are there signs/symptoms of catheter-associated urinary tract infection?

Monitor respiratory examination
Monitor ventilation with end tidal CO2 monitoring (if available) Monitor ventilation and oxygenation with arterial blood gas sampling (if available)

Monitor and adjust mechanical ventilation settings/modes
Are there signs/symptoms of ventilator-associated pneumonia?
Repeat chest X-ray if ventilator-associated pneumonia suspected Can removal of endotracheal tube be considered?

Neurological examination
Follow up neurosurgical consultation (if applicable) Level of consciousness (i.e., GCS, modified for infants) -is sedation required?
Has there been a change in examination since last review? '✓' can be selected when a proforma question has been answered, or a proforma point has been reviewed or tested. Temperature monitoring. Fever is associated with worse outcomes in neurocritical care and an increased one-year mortality in HIV-uninfected individuals with TBM 9 . Pyrexia may indicate superimposed bacterial infection.

Medication evaluation and management
Important characteristics to monitor for anti-TB chemotherapy are described below and in Table 2.
Anti-tuberculosis chemotherapy. The optimum delivery of essential anti-TB chemotherapy is a priority, but optimal doses and administration routes are unknown 18 . Prompt treatment and avoidance of therapy interruptions are essential to reduce mortality, and in unconscious patients, crushed medication administered via nasogastric tube, or intravenous therapy, may be considered 18,19 .
Anti-TB chemotherapy can change during the long duration of treatment. Regimen modifications may be necessitated by drug resistance, changes in weight, interactions with cytochrome P450-inhibiting anti-retroviral therapy (ART), and drug side effects, many of which cause and exacerbate critical illness 18 . Rifampicin, isoniazid, pyrazinamide and fluoroquinolones can cause liver injury; therefore, regular monitoring of alanine transaminase (ALT), aspartate aminotransferase (AST), and bilirubin is important. Optimal management of drug-induced liver injury in TBM is unknown and is currently being studied 20,21 .
Nutrition and the gastrointestinal tract TB causes a chronic catabolic illness and patients commonly present with weight loss or failure to thrive in children. Additionally, TBM patients are at risk of the negative consequences of the catabolic stress of critical illness 22 . Controlling glycaemia may be complicated by corticosteroids, which increase serum glucose. Nutrition in TBM often requires nasogastric tube placement when consciousness is reduced. Maintaining adequate nutrition is important to provide substrate for healing. Early feeding and avoidance of hypoglycaemia may improve outcomes after acute brain injury 23 , but has not been studied in TBM.
Multidrug regimens, corticosteroids and anti-platelet drugs can lead to gastrointestinal intolerance and bleeding 24 . Gastrointestinal bleeding may cause hypovolaemia and exacerbate reduced cerebral perfusion. Nausea and vomiting, common side effects of anti-TB chemotherapy, may further contribute to hypovolaemia. Aspiration of vomited gastric contents is a risk when consciousness is impaired. See Table 1 and Table 2.
Kidney function, fluid balance, and electrolytes Kidney function. Acute dehydration may place patients at risk for hypovolemia and prerenal acute kidney injury. Chronic use of anti-TB chemotherapy may be nephrotoxic 25,26 . Obtaining baseline kidney function tests, including creatinine and urea, may identify those at risk of acute or chronic kidney injury. Changes in urine output or fluid balance may precede laboratory abnormalities. See Table 1 and Table 2.
Hyponatraemia and fluid balance. Fluid balance is an important distinguishing parameter between the causes of hyponatremia (sodium <135mmol/L). Central venous pressure (CVP) reflects the intravascular volume and helps determine hydration status. CVP can be measured continuously or intermittently from a central venous catheter most often placed in the internal jugular vein or femoral vein if they are available. CSW is often characterised by high volume urine output, hypovolemia with low CVP, clinical signs of dehydration (dry mucous membranes, delayed capillary refill time, tachycardia, and hypotension) and haemoconcentrated laboratory parameters (elevated haematocrit, haemoglobin or urea) 27 . Conversely, the syndrome of inappropriate anti-diuretic hormone (SIADH) lacks a high volume urine output, and patients are usually euvolaemic with a normal CVP, and no clinical signs of dehydration 27 . Fluid balance charts may help identify these fluid shifts. Despite these distinguishing features, CSW and SIADH can be difficult to diagnose and other laboratory tests, such as serum and urinary osmolality and urinary sodium, may help further identify the aetiology, which is critical due to their divergent management approaches.

Hypokalaemia.
Hypokalaemia may be a result of drugs or of poor nutrition, whereas hyperkalaemia may be due to hypoadrenalism either directly from TB or from withdrawal of corticosteroids. Monitoring low serum potassium and replacement requires an environment capable of close monitoring. The use of hypertonic saline, fludrocortisone and acetazolamide, individually or in combination, may rapidly shift electrolytes.

Risks of prolonged critical care admission
Pressure ulcers are common in immobile individuals requiring prolonged care. Nosocomial infections occur due to changes in patients' immune system and placement of foreign objects, such as a central venous line, arterial line, urinary catheter or endotracheal tube. Deep vein thrombosis is a risk of prolonged critical care admission. Urinary tract infections may occur secondary to urinary catheters. Further, impaired consciousness may require prolonged intubation and mechanical ventilation 28 , which may be associated with a higher risk for gastrointestinal haemorrhage, septicaemia and pressures ulcers 29 . Prolonged recovery and slow ventilator wean may involve tracheostomy placement.

Neurocritical care
In addition to general and critical care management, specific neurocritical care may assist in TBM management.

Ventilation and oxygenation.
Oxygen administration may increase cerebral oxygenation and possibly reverse brain hypoxia 37 . Oxygen and carbon dioxide (CO 2 ) levels also affect CBF and ICP. Decreased arterial oxygen and increased CO 2 both dilate cerebral vessels, which may increase cerebral blood volume and ICP 38 . Conversely, aggressive hyperventilation may constrict cerebral vessels and cause ischaemia from decreased CBF 38 . Therefore, tight control of end tidal carbon dioxide (ETCO 2 ) is crucial to control raised ICP but avoid ischaemia and is an important parameter in TBI management guidelines, although the target level for TBM in unknown 34,39 . Both oxygenation and ventilation can be monitored noninvasively with pulse oximetry and ETCO 2 or invasively with arterial blood gases.
Temperature. Hyperthermia may increase cerebral metabolic rate and CBF, which can further increase ICP in a swollen brain 40 . Induced hypothermia is experimentally neuroprotective but is associated with poorer outcomes in TBI 34,35,39 . The target temperature in TBM is not established.

Head-of-bed elevation.
Elevating the head end of a bed lowers ICP through improved venous drainage and extracranial shift of cerebrospinal fluid (CSF). However, the MAP may also fall. In non-TBM pathology, studies suggest a beneficial or non-detrimental role of head-of-bed elevation to 30°4  . Repeat neuroimaging can be used to monitor disease progression, evaluate deteriorating patients, and confirm the placement of a ventriculoperitoneal (VP) shunt or an external ventricular drain (EVD). The ideal timing of follow up imaging is unclear in stable patients.

Raised intracranial pressure
Raised ICP is a key factor precipitating adverse outcomes. Firstly, it reduces CPP and exacerbates existing cerebral ischaemia due to vasculitis. Secondly, generalised or compartmentalised increased ICP causes brain shift and consequent neural injury. ICP monitoring is useful; non-invasive monitoring techniques may be used when gold-standard invasive monitoring is unavailable. New monitoring techniques have potential for improving patient care, yet these are not widely available. Important points for the monitoring and optimisation of ICP are shown in Table 3 and Table 4.

Non-invasive intracranial pressure measurement and monitoring Fundoscopy
Optic disc swelling can indicate raised ICP. However, fundoscopy is challenging to perform and highly operator dependent, and the development of papilloedema can be delayed.

Optic nerve sheath diameter ultrasound
Changes in optic nerve sheath diameter (ONSD) due to raised ICP occur rapidly and can be measured using ultrasound 58 .
ONS ultrasound is quick, easy, and reproducible, and correlates with ICP 59,60 , although evidence for its use in TBM is limited 61-63 .

Compromised cerebral perfusion
Various neuromonitors have been used in non-TBM pathologies, to detect ischaemic brain injury 64-66 . These measure various facets of brain perfusion and each have strengths and limitations.

Transcranial Doppler ultrasound
Transcranial Doppler (TCD) can be used to measure flow velocity in basal vessels and detect vasculopathy; however, it may not detect mild-moderate ICP changes, is limited to flow in the major cerebral vessels, and is technically challenging 18,67 .

Non-invasive cerebral oxygenation monitoring: near-infrared spectroscopy.
Near-infrared spectroscopy (NIRS) is a non-invasive monitor that uses optical technology to continuously assess brain oxygenation 68,69 . NIRS is limited by superficial penetration of cortex, distortion by the skull, CSF and oedema 68,70,71 and poor long term monitoring.
Invasive cerebral oxygenation monitoring: partial pressure of brain tissue oxygen tension. The partial pressure of brain tissue oxygen tension (PbtO2) monitor is a thin parenchymal catheter that offers continuous monitoring of brain oxygenation 66,72 . Normal values have not been established; however, the risk of poor outcome increases with PbtO2 <20mmHg 73 , especially <10mmHg 64,72,74,75 .

Invasive intracranial pressure measurement and monitoring.
CSF opening pressure may be measured from the ventricles with an EVD or via lumbar puncture (when safe) 76 . CSF drainage allows simultaneous ICP monitoring and treatment. Continuous monitoring is possible with a parenchymal probe.
Post-operative neurosurgery management. This includes wound review, suture removal, and clinical monitoring for signs of treatment failure. Repeat imaging can check treatment success. EVDs must be carefully managed to avoid lifethreatening complications (infection and overdrainage-related intracranial haemorrhage). VP shunts are permanent and therefore complication rates must be viewed over the full lifetime of the patient.

Management of acutely decompensating patients
Causes for acute neurological decompensation include raised ICP, metabolic disturbances (i.e., hyponatremia, hypoglycaemia), stroke (ischaemic or haemorrhagic) and seizures. Table 4 outlines a priorities-based checklist approach to the acutely decompensating patient. The rationale for this checklist is described below, unless already discussed.

Seizures
Clinical and subclinical seizures can increase ICP due to the increased cerebral metabolic demand and resultant increased CBF. Hydrocephalus, infarcts, tuberculomas, and electrolyte imbalance can all precipitate seizures. Anti-convulsants that induce cytochrome P450 enzymes, or are susceptible to enzyme induction by rifampicin, may complicate management.

Hyperosmolar treatment
Intravenous administration of a hyperosmolar solution creates an osmotic gradient, removing water from the brain and decreasing ICP 77 . Hypertonic saline may lower ICP faster, further, and for longer than mannitol 18 ; however, no trials have directly compared these agents in TBM.

Raised intracranial pressure surgical management
Hydrocephalus VP shunting has long been standard practice for hydrocephalus but may be associated with complications 78,79 . EVD may be used for temporary drainage of CSF, and to assess the benefit of a VP shunt in patients with an altered sensorium. With endoscopic third ventriculostomy (ETV), CSF is drained internally by connecting the ventricles with the subarachnoid space via a stoma in the floor of the third ventricle. ETV is particularly challenging in TBM and experience is required 80,81 .

Mass lesions
Surgical excision for tuberculomas is uncommon but may be indicated depending on their size, location, expansion, and clinical consequences. Surgery is more commonly needed for TB abscesses (drainage and/or excision).

Cerebral venous thrombosis
Cerebral venous thrombosis is an unusual cause of acute neurological deterioration in TBM, but has been described 82,83 .

Data availability
No data are associated with this article.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
Author 1. The authors have attempted to streamline the management of TBM and its complications in a step wise fashion and is a welcome step. However, there is nothing new in what they have mentioned and is practiced in the countries where the disease is endemic. I am fine with summary for management except for a small comment on table I and is as under: Table 1: Assessment for communicating hydrocephalus with air encephalogram or column test is obsolete and is only of historical relevance and should be deleted.
Thank you for your comment. However, we disagree that this does not offer anything new; our proforma is the first comprehensive patient assessment tool for tuberculous meningitis. No other checklist exists to allow a priority-based approach to a deteriorating patient in this disease. Whilst the information included in the checklist and proforma may be accepted knowledge, or available elsewhere, our presentation of this information in proforma and checklist formats aim to support clinical assessment, and highlight vital components, during clinical care.
2. Table 1: Assessment for communicating hydrocephalus with air encephalogram or column test is obsolete and is only of historical relevance and should be deleted. We note, but disagree with, the reviewer's comment on air encephalography and column tests. These are used as standard approaches in at least two big centres that publish on TB meningitis, based on published data, so this is hardly historical nor obsolete. There is no current technology, apart from invasive methods, that have been shown to safely and reliably distinguish between communicating and non-communicating hydrocephalus. This is based on published studies. We know that some centres do not try to distinguish and therefore have higher rates of surgical procedures -VP shunting and endoscopy. However, with medical management, most patients can avoid those surgical procedures -this was published by Johan Schoeman many years ago and the results are as relevant today as they were then. But this of course depends on being able to do lumbar punctures safely, which may be risky for the 15-20% of patients that may have non-communicating hydrocephalus. To our knowledge, there has been no paper showing the safety and reliability of any imaging to confirm the communicating nature of hydrocephalus in TBM. So the reviewer's comment is not evidence-based and we are comfortable that our manuscript reflects published data.
No competing interests were disclosed. Competing Interests: 03  Still under temperature monitoring. The reader is interested in clues as to how to differentiate the temperature of TBM from super imposed infection? Table 1 page 4: Relevance of the previous BCG scar? I propose that the tests be separated in to those with high utility like gene expert, low CSF glucose from those with low utility like AAFB, culture. Imaging must always be contrasted unless contraindicated   I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. 7. Imaging must always be contrasted unless contraindicated This is a good point. However the addition of contrast may not always be possible. We have added 'preferably with contrast' to table 4 8. Table 2: The part on monitoring for GIT bleeding, what is the practical way of doing this? Not so much information about its relevance in TBM patients. Recommending how to practically monitor for GI bleeding goes beyond the scope of this article. This adverse event is relevant given frequent dexamethasone use in TBM, and the increasing evidence base for aspirin 9. If possible the authors should expound on how to suspect superimposed infections Although this is an important component to TBM care, we feel expanding on this is beyond the scope of this article 10. Table 4  We have now specifically mentioned IRIS and cerebral sinus thrombosis in table 4. In the supporting document we have added the sentence 'Cerebral venous thrombosis is an unusual cause of acute neurological deterioration in TBM, but has been described' with references.
11. Risks of prolonged hospital admission should include DVT. We have added 'Deep vein thrombosis is a risk of prolonged critical care admission' 12. Page 9 Last paragraph under neuroimaging needs a reference. We have discussed the repeat imaging with references in the preceding paragraphs, and repeat imaging after placing hardware in the brain is standard neurosurgical practice No competing interests were disclosed. Competing Interests: