Healthcare: What’s the price of staying alive?

Random sampling of the news in recent times about rising healthcare costs:

https://www.straitstimes.com/singapore/hard-choices-need-to-be-made-for-sustainable-healthcare-pm
https://sbr.com.sg/healthcare/news/medical-inflation-in-singapore-hit-10-in-2019

And yet in the real private sector, prices of long time products have been on the decline:

https://www.ikeahackers.net/2018/02/10-new-lower-price-items-hack.html
https://blog.seedly.sg/how-much-have-mcdonalds-prices-increased-over-the-years/

So the immediate question that can be asked: Should the healthcare sector just hire folks from Ikea and Mcdonalds to run our healthcare facilities? From a pure costs perspective, that might not be such a bad idea. But from a quality perspective, it seems to have remained the same for Ikea and Mcdonald’s all these years.

So what are some of the strategies they have employed and is there a viable solution to ballooning healthcare costs?


Lean out production process

Both Ikea and Mcdonald’s has a heavy focus on lean for production. If you recall close to a decade ago, there might be cold burgers that were mass produced just waiting to be ordered and you would choose the burger that is the freshest despite not being your first choice. Fast forward probably 5 years ago, Macs revamped their outlets and created a just in time demand pull system across all their outlets. This ensured all burgers and fries are fresh and hot when they are served.

To my knowledge most hospitals already practice lean and treat patient flow like a production line. I think the key issue is that the flow is not optimized with IT and is still too disjointed. But due to the crappy IT ecosystem that has evolved (discussed later).

Self serve

Ikea drives down prices with self serve model. You pick up your own furniture, you deliver your own furniture, you assemble your own furniture.

There is minimal self serve such as visitor registration and payments. So far these have been good but limited in impact.

I’m not quite sure if this can be translated to acute care services. But for step down care and pharmaceutical care this could work. If Japan has a dispenser for soiled panties, I am sure a dispenser for medication is equally viable but I’ll return to this point (with the discussion of electronic medical records held by patients)

Expanded portfolio of products

Both companies continues to innovate with new products that move up the quality chain and charge customers a premium. A decade ago I would never expect myself to pay more than $8 for a McDonald burger but fast forward to today, a lot of my friends enjoy the burger despite the relatively higher costs. Perhaps with inflation things other prices have caught up, but I think McDonalds has grown their portfolio to have a larger range of items with higher revenue drivers. The range of portfolio and profit margins help to maintain a profitable business.

Healthcare on the other hand might suffer from a different issue where there are too many products. Just implants alone there could be in the range of hundreds just for hip replacement (material, size, technology, manufacturer etc). So perhaps for healthcare a reverse could happen where the items could be grouped into bundles like telecom bundles to help reduce cash being held up by inventory. That or at least for implants hopefully 3D printing technology catches up so that there is price parity between mass manufactured items and bespoke 3D printed items.


IT to the rescue

Close to 20 years ago, Singapore embarks on digitizing their hospital operations:

These 5 pilot applications focus on the processing of claims in hospital bills, electronic procurement of medical and surgical supplies, the transmission of notifications to government agencies, a national patient master index and access to local and foreign databases. MediNet will be able to link the computer systems of all participants in the healthcare delivery system. These include the Ministry of Health, Central Provident Fund (CPF) Board, government, restructured and private hospitals and general practice clinics. By connecting their computers (personal computers, minicomputers and mainframes) to MediNet, these organisations will be able to communicate, exchange data and access common applications on the network. — https://www.ncbi.nlm.nih.gov/pubmed/2260820

This was a good first step to improving productivity and retaining data in a digital form. By the time I joined healthcare 4 years ago, the application was mostly run by heavily customized off the shelf solutions (I know it sounds ironic) such as SAP Industry Standard Healthcare, Epic, Cerner diagnostics. The last bastion of bespoke software was CPSS that was used for dispensing drugs at the pharmacy.

From my perspective, this is a huge drain of resources with millions of dollars spent annually maintaining crappy antiquated systems likely to breakdown at any time. And with the Singhealth hack (https://www.straitstimes.com/singapore/personal-info-of-15m-singhealth-patients-including-pm-lee-stolen-in-singapores-most) this old, clunky and ugly systems were not really as secure as they professed to be.

So why not move to a more modern system? For starters, the sector is run by risk adverse bureaucrats. Not necessarily a bad thing for stability, but definitely not a good thing for innovation. And if vendors have been leeches sucking the life out of you for the last 10 years, change is likely needed before there’s no more blood left to be drained.

The hierarchy looks something like this

MOH looks after policy

MOHH is the holdings group to handle profit and loss for the layers underneath (this is a guess on my part)

IHIS looks after the IT infrastructure and applications (essentially from an org structure they are the same as the hospitals and facilities but in terms of IT they look after the medical facilities)

So far the government has managed to transform parts of the bureaucracy with GovTech. The new spin off serves as an in house consultancy to help bring design and technology to the civil service.

I strongly feel that IHIS should be transformed to become the GovTech for healthcare. That would allow the hospitals to transform digitally with little or no reliance or consultancies and development houses. A possible strategy could be to seed folks from GovTech and grow the mindset and skillsets of the healthcare IT arm.

Once the talent pool has been trained properly, I would think the immediate area to tackle is to build a proper electronic medical record with the patient at controlling their own medical data. Right now, all medical data resides with medical facilities. If you visit a hospital (Tan Tock Seng) and visit a different hospital cluster (such as Changi) you’d likely have to fill in your details again. They would have no records of your medical history without having to call up the other hospital. With the internet being used to spread cat memes, it baffles me how we are not using broadband and development efforts for something more meaningful. And if the medical records were designed with the patient being in control, they can carry it any where they need to. A possible design is also to have doctors/ medical practitioners unlock together as a two key lock system. (This prevents patients from committing insurance fraud) every entry will be signed off by a medical professional and then locked by the patient (and no, we do not freaking need blockchain for this crap). Prescriptions could be unlocked at the pharmacy by the pharmacist.

With the backbone of the medical records being built, other applications can be bolted on such as health records, drugs record, diagnostics etc. This helps to build a more comprehensive picture for the individual.

The pricing engine that is the main contention of migration will need to be rebuilt from scratch. The SAP FICO backend could still be viable (It is a good system) but definitely do away with the patient accounting module. This should be redesigned with MOH, MOHH, IHIS, CPF and Insurance all as stakeholders. It will take time, but healthcare is complicated so let’s get it right (patients should not need to worry about the payment. MOH should figure out at the policy level on how to take care of the citizens and work with the appropriate parties to execute their policies)

Presently healthhub attempts to do consolidate records, but it’s just way too complicated and poorly designed so it has poor scalability. The design debt with the ridiculously long hidden menu is just bad UX. The integration with SAP Patient Management for appointment booking is a good but expensive attempt at building on top of rotting foundations.

I do hope that efforts can be re concentrated to build proper products for the healthcare scene that can hopefully lower costs in the long run


Let’s revisit this long gripe:

  1. Talent reskilling — Create a GovTech for healthcare with IHIS folks
  2. Build electronic medical record, replace the existing licensed crap from SAP/ Epic
  3. Extend and replace functionalities of existing applications with applications built and maintained in house
  4. Lessons from the private sector such as creating self serve flows can be further explored
  5. Lean out the process further