At night the Ebola hospital looks like the circus has come to Moyamba. There is little electricity in the town and our illuminated tents and security lighting provide a festival of light.

The night shift has a skeleton staff, but with only two patients the staff-to-patient ratio would be the envy of the NHS. Few patients are admitted during the night in rural Africa - road traffic accidents and obstetric emergencies mostly.

On my first night on call I get just one suspect case referred. Unfortunately, the case is a 17-year-old who is pregnant. Pregnancy and childbirth is at the highest risk for transmission due to the potential for exposure to viral-rich blood and amniotic fluid during labour.

The maxim goes: don't undertake any procedure where you can’t see your fingers, which essentially means midwifery becomes a spectator event. The only intervention is sublingual Misoprostol after delivery to minimise postpartum bleeding. Even cutting the cord is viewed as a risky procedure. The baby always dies and the mother usually follows.

Since it is over 20 years since my last obstetric experience, this does not seem like the ideal time for a quick refresher. Fortunately, the ambulance makes a mistake and takes her to the government hospital rather than to us. I breath a deep sigh of relief.

We are ghost-like apparitions in our white biohazard suits under the fluorescent lights of the walkways.

Saffie has deteriorated after her brief resurgence this morning and I think she is going to die overnight. I am dejected. We have arrived with fanfare and hope and we are failing. We are doing everything we can, but we are losing. We only have two patients, but they are an early test from the community about the difference we will make.

During the night I try to catch glimpses of her from the green zone and return with hope when I see her lift an arm or move her head.

We have 100 beds in the hospital (more realistically 60-70 to avoid the beds being too close to each other in the tents) and earlier in the day we have been discussing how quickly to scale up. The medics are keen to expand rapidly (quelle surprise) but the water/sanitation (watsan) lead is as usual far more cautious and wants to increase capacity by three beds per week.

We point out that the epidemic will be over by the time we reach full capacity, but we know from the last week that this is not an argument that we will win (we never win any arguments with them). Chris manages to negotiate a peace pact whereby the clinical teams can go in solo without a murmuration of chlorine sprayers swooping around us all the time. This is a real breakthrough.

I am surprised there have not been more suspected or confirmed cases referred to the hospital during the day. The numbers recorded at the Command and Control meetings do seem to be falling in recent days and my next anxiety (my anxieties pass like batons in a relay race) is that we won’t have enough patients to fill our empty new beds.

I speculate four hypotheses for the unexpected low numbers in recent days:

1) The epidemic is over. This is my favourite, but probably a little optimistic. While it is my favourite, it also sparks a flicker of disappointment. Having put so much into setting the centre up, it feels a bit of a shame not to be able to use it fully. This feeling troubles me as I think I might be Dr Evil.

2) The rapid increase in UK-funded Ebola Centre beds in the last couple of weeks is dwarfing demand. The number of beds has doubled from 500 to 1,000, and it may be a basic market forces issue: excess supply and diminishing demand.

3) People are hiding their sick relatives. We know this is happening in a number of chiefdoms, especially with Christmas coming. Ribbi in particular, where I will go tomorrow to investigate further.

4) Ebola is taking a bit of a break for Christmas. Perhaps going home to er, Ebola, for the holidays.

Hypothesis 1) is not as far-fetched as it might appear. I hear that the Ebola Centres in Bo and Kenema have been contacting the Ebola response teams to ask for patients, as they have seen patient numbers fall in the last week. Kailahun, the first centre that experienced the Ebola epidemic in Sierra Leone, has apparently discharged its last patient. So it may be that the tide is finally rapidly turning.

Hypothesis 2) has validity as the rapid increase in capacity will be spreading demand. I envisage the new Ebola centres fighting for business in this new competitive environment. I join the fray and start emailing all my contacts in Freetown and national centres to advertise that we are open for business. Maybe some TV advertising.

Hypothesis 3) is also a real concern. House-to-house searching has begun in Freetown, and tomorrow it begins in Moyamba. The local Chief has decreed that the search will be for the sick, the dead and strangers who are not allowed to stay at home. The sick and the dead will be reported (I’m not sure this threat will work for the dead). We should expect to uncover more cases.

The next few days will be test these hypotheses. On the dawn ward round Saffie is alive and stronger. I dare to hope. Meanwhile, rather sleepily, I am off to confront the rebels of Ribbi.

Our efforts to look after them are hindered by language barriers - none of them speak Mende and our nurses speak little Timini or Shabu, the languages from their home district of Ribbi.

MORE BLOG POSTS FROM PROFESSOR JOHN WRIGHT