8.4 Analysis and evaluation at the cholera facility level

8.4.1 Weekly number of admissions

The number of cases admitted each week gives an indication of the evolution of the epidemic and the intensity of transmission in the catchment area of the facility.
The follow-up of weekly number of cases also guides facility management in terms of treatment supplies, beds, and human resource needed.
Early in the outbreak, if the trend in weekly admissions shows that the number of expected cases was underestimated, a prompt adjustment of supply stocks, bed capacity and human resources should be undertaken.
Late in the outbreak, as the admissions begin to diminish, a decision can be made to close a certain number of beds and assign staff elsewhere.

8.4.2 Weekly case fatality rate

At the start and end of an outbreak, the CFR may be higher than 1%, as there is a relatively small number of patients and one single death represents a high proportion of cases admitted.
At the start of an outbreak, the CFR may be higher in a CTC for 1 to 2 weeks if the overall functioning of a CTC is not yet optimal.

Note: there is evidence that the overall reported cholera mortality during an epidemic can be significantly underestimated when only treatment facility deaths are counted1. Adding community cases and deaths to facility data gives a more representative view of the overall situation.
At the very least, high CFR is associated with inadequate treatment access band/or public information about available care (facility location, gratuity of services, etc.).

8.4.3 Age group

During an outbreak in a non-endemic setting, all age groups are equally at risk for symptomatic cholera. Thus, the proportion of patients in each age group is similar to the overall representation of this age group in the population (i.e., for children < 5 years of age: 17 to 20%).

During an outbreak in an endemic setting, natural immunity accrues as people are repeatedly exposed to Vibrio cholerae over time and cholera cases in the < 5 age group are proportionally more common (roughly 25 to 35%).
Toward the end of an outbreak, as the overall number of cases declines, the proportion of cases in children < 5 years of age admitted typically increases (e.g., to above 50%) because watery diarrhoea in children from common non-cholera causes again begins to predominate.
If the proportion of children under 5 years is much less than expected, look for problems of access to care for this age group.

8.4.4 Geographic origin of patients

It is recommended to register and analyse geographic origin of patients so as to follow the evolution of the epidemic and allocate resources to the populations most affected.

A concentration of cases in a particular location suggests that there is a common source for cholera infection (i.e., river or contaminated well). Once the source is identified, appropriate control measures can be implemented.

If the analysis of patient origin shows that populations are being affected successively along a river or transportation route, prevention measures may be implemented to protect those that are not yet affected.

An increasing number of patients coming from outside the catchment area for a given CTC may indicate that the current cholera facility should be relocated or that additional treatment site(s) should be established in the area where these patients live.

8.4.5 Sex

Biologically, men and women are equally susceptible to cholera infection and symptomatic disease.

At the onset of an outbreak, men or women may be more affected due their occupation. This provides information on the source and/or mode of transmission of Vibrio cholerae. However, as the epidemic evolves, the number of men and women tends to equalize.

If the proportion of female adults and children is much less than that of males, look for problems of access to care for this group.

8.4.6 Cholera vaccine status

If there has been cholera vaccination in the years preceding the current outbreak, recording the vaccination status of patients admitted in a treatment facility helps investigators and public health officials in evaluating the effectiveness of the vaccine and vaccine strategy.

8.4.7 Pregnancy

Pregnant women normally represent 2 to 6% of the general population and should represent a similar proportion of the admissions to a CTC.

8.4.8 Circumstances of patient death

A small number of patients who arrive at a CTC/CTU with hypovolaemic shock or a significant comorbidity may die even with appropriate management. The vast majority of patients, even the most severely dehydrated or fragile should be successfully treated.

For all death
– Systematically review the patient file, taking into account time since admission (see below).
– Look for erroneous or inadequate:
• diagnosis
• prescription
• procedures
• patient surveillance
Specifically:
• Look for missed diagnosis or mismanagement of concomitant severe acute infections (e.g. severe malaria) or cholera complications (e.g. hypokalaemia, hypoglycaemia).
• Check if individual risk factors (e.g. extreme of age, severe acute malnutrition, or known cardiovascular disease) were taken into account.
• Determine if death directly results from a specific error in care (e.g. fluid overload, rapid administration of parenteral potassium, airway obstruction from incorrect nasogastric tube placement) or from an overall lack of proper care (untreated dehydration).
– Check for a lack of essential medical supplies, personnel, technical competencies, etc.

Deaths within the first 4 hours of admission
As for all death (see above) and in addition:
– Look specifically for:
• inappropriate waiting time in triage;
• inadequate critical care management (e.g. hypovolaemic shock).
– If the patient was transferred from an ORP, determine if the duration of transport and/or a lack of care during transportation played a role in deterioration of the patient’s condition.
– If patients arrived by their own means, check patient geographic origin with respect to the location of the treatment facility. Distance, lack of transportation, and/or insufficient public information on how to seek care, can result in critical delays in treatment.

According to the problems identified, the following measures should be taken:
– Reinforce training and technical skills, particularly where severe patients are treated.
– Ensure that sufficient staff and medical supplies are present day and night, particularly where severe patients are treated.
– Ensure that protocols are properly displayed in all wards.

In addition, for deaths within the first 4 hours of admission:
– At CTC level:
• Assign staff experienced in rapid resuscitation to triage.
• Review patient transfer procedures and care prior to or during transfer.
– Outside the CTC:
• Consider establishing an ORP or even CTU in distant location from where new patients are beginning to arrive. Consider home-based treatment when the above option is not feasible.
• Reinforce health promotion activities.

8.4.9 Dehydration level on admission

CTC

In a “referral” CTC, i.e. a CTC that receives patients referred by ORPs or CTUs, cases with severe dehydration can represent 70 to 80% of the total number of cases.

If the CTC is not a referral CTC, but a local CTC receiving all types of patients, the distribution of cases with the most common circulating cholera strain should be the following: approximately 25 to 30% of cases with severe dehydration, 30 to 40% with some dehydration and 30 to 40% with no dehydration.
If proportions differ greatly from these averages, check the case management and/or response system set up e.g.:
If the proportion of “no dehydration” is higher than expected, ORPs are needed to free up the CTC.
If over 30% of patients are diagnosed with “severe dehydration” on admission:
– Check diagnostic accuracy in the CTC:
• Staff may not be sufficiently trained and therefore over diagnose the degree of dehydration.
• Dehydration may be overestimated to justify IV treatment that is perceived as more effective. Overuse of IV treatment exposes patient to complications, lengthens hospital stay and increases the cost of treatment.
In this event, reinforce staff training and competencies.
– If patients have been referred by an ORP: check triage effectiveness and the accuracy of prescriptions at ORP level, as well as the transfer system (delay due to dysfunction e.g., vehicle breakdown, communication means).
− Check the geographical origin of patients and the rate of severe dehydration in those that come from a zone that is not covered by the CTC. Individuals may have to travel long distances to receive treatment and their dehydration may get worse on the way. In this event, consider setting up an ORP or even a CTU in the distant area where the patients are coming from; consider home-based treatment if it is not possible to set up an ORP.

ORP

When a CTC/CTU exists in the affected area, ORPs should have few cases of severe dehydration.
If this is not the case:
– Check diagnostic accuracy. If there is over-diagnosis of severe dehydration, reinforce staff training and competencies.
– If diagnosis is correct, a high number of cases of severe dehydration indicates at the very least people do not seek medical care enough. Analyse the causes (CTC too far away, cost of transport, etc.) and try to improve access. Reinforce health promotion activities.

8.4.10 Length of stay

The average length of stay in a CTC is 2 to 3 days. Usually, a patient without dehydration remains under observation a few hours. A patient with some dehydration and no complications stays in general one day and a patient with severe dehydration or complications may remain hospitalised 4 to 5 days.
If average length of stay exceeds 3 days, review case management. Insure that prescriptions and surveillance are adequate. If patient monitoring does not track on-going fluid loss, prolonged dehydration or cycling between hydration and dehydration occurs, resulting in increased length of stay.
Also check that patients are not mistakenly kept hospitalised until diarrhoea has resolved1 .
The length of stay can be longer if the patient has serious comorbidity. When certain comorbidities are frequent, e.g. numerous cases of acute malnutrition among children under 5 years, consider other interventions such as implementation of therapeutic feeding centre.

8.4.11 Leaving against medical advice

The number of patients who leave prior to the completion of their rehydration therapy should be very small. The reason for leaving should be for personal reasons only (e.g. unattended children at home or fear of losing their job).
If there are more than a few patients leaving against medical advice, assess the conditions in the treatment facility. Physical discomfort (e.g. extremes of temperature, mosquito or fly infestation), unpleasant staff, lack of information on the disease and the necessary treatment or overall facility dysfunction may provoke patient departure before completion of therapy.

8.4.12 Consumption of RL and ORS

On average, a patient needs 8 to 10 sachets of ORS and a patient on IV treatment needs, in addition, 8 to 10 litres of RL.
By calculating the average real consumption of ORS and RL per patient, it is possible to estimate how many patients can be treated with available stock and how long available stock will last.
Furthermore, analysis of consumption identifies possible over or under consumption of ORS and RL.

In general, under-consumption usually concerns ORS and over-consumption RL. In both cases it is important to identify the reasons. For example:

The number of ORS sachets per patient is under 6
– Check for possible prescription errors (volume of ORS prescribed insufficient).
– Check if IV treatment is incorrectly prescribed in patients that should receive oral treatment.
– Compare the volume of ORS prescribed and the quantity actually drunk, to check whether there is a problem in the monitoring of patients.
– Check that prepared solution is available at all times on the ward.
– Check patients on IV treatment are drinking ORS (early introduction of ORS to patients on IV treatment is often neglected).
– Check that discharged patients are leaving with 2 to 4 ORS sachets to take home.
– Check that preparers use one sachet of ORS per litre of water, that sachets are not being overdiluted.

The number of litres of RL in patients on IV treatment is over 12
– Check that IV treatment is not administered too long in patients that could come off IV treatment and switch to oral treatment.



Footnotes
Ref Notes
1

The goal of hospitalisation is to ensure patient rehydration until vomiting ceases, the diarrhoea has clearly decreased and that the patients can safely treat themselves at home until diarrhoea resolves (Section 5.1.10).