Care bundle intervention

The clinical care bundle intervention tested in this study was designed using two

concomitant approaches. Firstly, all published data and guidelines on clinical care bundles in the literature for infectious diseases, critical care and emergency medicine, plus general emergency resuscitation guidelines for acute medical and surgical emergencies were reviewed. The details of these searches can be found in Chapter 2 Section 2.1. The data from the analysis of patients with meningitis in Malawi to determine which predictors of poor outcome would be amenable to treatment with a care bundle was additionally reviewed (Results Chapter 4 Section 4.5.3-4). The data from this analysis was combined with the published data and guidelines to create a simple clinical care bundle that would meet the resuscitation needs of the screened patients within the resource constraints of the clinical environment in Malawi. The care bundle was additionally designed to be a nurse-led

120 intervention that enables a nurse or clinical officer following a detailed protocol to identify physiological abnormalities from baseline observations and use the protocol to correct these safely over a 6 hour time period. The protocol gives clear clinical targets that should be achieved by 6 hours where possible.

At the end of the 6 hour time period, every effort should have been made to optimise the clinical care given with the bundle to normalise physiological abnormalities and stabilise the patient prior to ward transfer.

3.3.7.1 Published guidelines on the approach to critically ill medical patients

The clinical care bundle of early goal directed therapy tested in this study was designed primarily based on the ‘ABCD’ approach to resuscitation which has been developed and validated by the American and UK resuscitation councils (Council, 2010). These guidelines indicate the priority order of emergency life-saving interventions that should be given, according to clinical need to patients presenting with acute illness. The most urgent is

considered the airway “A”, followed by breathing “B” where oxygenation must be optimised, and are followed by “C” circulation and “D” disability. The ABCD system however is designed

primarily for patients in cardiac arrest and not presenting with a severe medical condition such as bacterial meningitis. As discussed in chapter 2, the components of the care bundle advocated by the Surviving Sepsis campaign were revised, and those which were applicable to the needs of the Malawian patient with ABM were determined, based on the clinical predictors of poor outcome from the previous analysis.

3.3.7.2 Clinical predictors of poor outcome from bacterial meningitis in Malawi

The results of the analysis detailed in section 1.5 of this chapter were reviewed. The detailed results are presented in Chapter 4 Section 4.5, however coma, altered mental status,

121 each individual predictor of poor outcome, the care bundles from the ABCD and the

surviving sepsis guidelines were reviewed, along with the literature from appropriate parallel specialities such as neurosurgery and adapted for a resource limited environment. The literature supporting each care bundle element is discussed in detail in Chapter 2, Section 2.8. The care bundle tested was designed to optimise clinical care for each predictor with the aim to restore normal physiology within a 6 hour observation period.

3.3.7.3 Final care bundle

The care bundle consisted of eight components, all of which were clinically available in the AETC and all of which are recognised existing clinical treatments which have supportive evidence of efficacy. The individual components of the bundle can be seen in Table 3.2 below, the ABCD order as determined by the Resuscitation Council guidelines.

A

Airway support is deemed essential by all guidelines for adults with a GCS of less than 8/15 at which point airway reflexes are minimal. Altered mental state is associated with raised intracranial pressure in meningitis, and the head tilt is designed to reduce ICP while optimising oxygenation and minimising aspiration risk, as shown in the neurosurgical literature.

B

Anaemia and hypoxaemia were identified as predictors of poor outcome; these were addressed in the care bundle by the application of oxygen via a concentrator for hypoxic patients, and the rapid transfusion of blood where available to those who were determined to be profoundly anaemic on a Hemocue test.

C

Tachycardia was shown to be independently associated with mortality; very few patients were hypotensive in the severity analysis. However hypotension leads to under-perfusion of cerebral tissues and therefore fluid resuscitation was given to patients with a clinical

122 definition of shock as determined by the surviving sepsis campaign (Table 3.4) to optimise cerebral perfusion.

D

Seizures at presentation and prior to admission to hospital were associated with poor outcome in the severity analysis, and are associated with an altered mental state. Therefore acute treatment of seizures and the underlying causes of seizures such as hypoxia and hypoglycaemia were addressed in the care bundle using intravenous glucose and facial oxygen where indicated.

In addition, early antibiotic therapy is associated with improved outcomes from meningitis elsewhere in the world, and in Malawi, trends towards higher mortality from meningitis when antibiotic treatment is delayed were observed in a small number of patients in the pilot cohort. We attempted to give the first dose of ceftriaxone within one hour of arrival in AETC in this care bundle.

Table 3.2 Clinical components of the care bundle Bundle components

1. Naso-pharyngeal airway if GCS <8 2. Head up/bed tilt 30° if GCS <11 3. IV access, 2g ceftriaxone stat

4. Oxygen via concentrator if SpO2 <93%

5. IV fluid (Ringer’s Lactate) bolus 20ml/kg if clinical shock, 125ml/hr if no clinical shock (Table 3.4)

6. Transfusion of packed red cells if haemoglobin <6.0g/dL

7. Correction of hypoglycaemia with oral dextrose and drinks/food if GCS >11, or with iv dextrose 10-50% via a bolus/infusion if GCS <11

123 This was due to difficulties in assessing staff training when no formalised resuscitation training was available that is appropriate to resource limited settings. During phase 2a, 6 weeks of time were devoted to care bundle training. The senior study nurse led the training of the other nurses at the bedside and in a series of practical seminars. Each of 6 weeks were devoted to one particular bundle element, and staff debriefed after recruiting a patient and following that protocol to discuss how to optimise the delivery of each care bundle element. During this time active study patient recruitment continued with a target of one patient per day. The study team recruited that patient as a team and delivered the care bundle intervention as dictated by the study protocol. The PI was involved in the initial staff training and in the daily de-brief meetings where what had been done well and which targets had not been met were discussed. The team then decided each day how to take the training forward to the next day. By the end of the 6 week period the team were entirely comfortable to deliver the care bundle to an individual patient without the support of their peers, but with 24 access to the PI by telephone.

3.3.7.4 Care bundle delivery and timing

The intervention care bundle was delivered using defined targets as detailed in Table 3.3 within the AETC. The duration of the intervention was 6 hours from initial screening, by which point all attempts to meet all targets using the protocols had been undertaken. Each subject received hourly observations including Glasgow Coma Score (GCS), blood pressure and pulse measurements. A lumbar puncture for the diagnosis of ABM was performed at the earliest opportunity. After each hourly observation, the study nurse reviewed the care bundle progress and added or stopped elements as per protocol. For example if the GCS dropped below 11 at hour 4, a head tilt was given.

124 Table 3.3 Targets for EGDT in BAM

Clinical targets for the care bundle intervention

Parameter Target Intervention Test of

intervention Timing of clinical assessment Medical review <1 hour of arrival Training in recognition of the symptoms and signs of meningitis and rapid triage

Timed and signed flow chart Antibiotic therapy 1st _{dose within 1}

hour of arrival

Education of importance of antibiotics, sepsis flow chart

Timed drug chart

Glucose BM >4 S/L , NG or IV dextrose Repeat BM

Oxygenation Sp02 >94% Nasal flow 02 from concentrator

Packed red cell Transfusion if significant anaemia (Hb 6.0g/dL) Regular Sp02 monitoring Perfusion CRT<2 sec, BP >90 syst, MAP >70, no postural hypotension (lying-sitting)/ UOP>0.5ml/kg (if catheterised IV fluid bolus 20ml/kg Ringer’s lactate where available. Repeated fluid balance assessment at 1hour then repeated

Seizures No seizures Acute prompt treatment of seizures with IV/PR benzodiazepines and IV phenytoin if seizures persist

125 At the end of the 6 hour study period, the bundle was stopped, and the subject was

transferred to the medical ward for ongoing routine care as deemed appropriate by the admitting clinician. A flowchart of the study process is available in the appendix. Patients were discharged to the most appropriate clinical care area in both phases of the study. If invasive ventilation or inotropic support was required, they were transferred to intensive care. Patients requiring on-going oxygen or control of persistent seizures they were transferred to the high dependency unit on the medical ward. All other patients were transferred to the main male or female medical wards.

Clinical shock was defined by abnormalities in the following general variables defined in Table 3.4. If shock was present then the fluid bolus protocol was instituted as per protocol (appendix 1.3). All protocols are in the appendix .

Table 3.4 Definitions of shock from the surviving sepsis guidelines 2008 Organ variables for the definition of clinical shock (Dellinger et al., 2008b) General

variables

Hemodynamic Organ dysfunction Tissue perfusion

Heart rate >100 bpm

Arterial

hypotension (SBP <90 mm Hg

Acute oliguria* Decreased

capillary refill (>3 seconds)

Tachypnoea RR>25

MAP <70 mm Hg Ileus (absent bowel sounds) Skin mottling

Altered mental status

SBP decrease >40 mm Hg

Blood lactate >4 mmol/L

*(urine output <0.5 mL/Kg/hr or <45 ml/hr for at least 2 hrs (if catheterised)/ no urine production for 12 hours despite adequate fluid resuscitation)

126 3.3.7.5 Staff training

Staff training formed an important component of bundle delivery during Phase 2a and was on-going through Phase 2b, but did not form a specific bundle element. This was due to difficulties in assessing and including the effects of staff training in the data analysis.