Thirty-two years on both sides of the hospital contract

Mandate

Keep clinical and administrative work running through system failure and change, at a cost the group's narrow margin can bear.

Outcomes

• Moved three hospitals onto one standard electronic record by November 2024, reducing customisations from 286 to 41
• Chose ¥84 million of isolated backups over a ¥238 million replica and rehearsed a 58-hour full restore
• Kept reception staffed through the insurance card transition while returned eligibility claims fell from 1,340 to 520 a year
• Introduced monthly reporting of doctors' after-shift time in the record

Decisions in context

The chairman created this post in 2020 to bring information systems into the executive group rather than leave them inside administration. I report to him, attend board meetings as an executive officer and sit on the group's medical safety committee. My work covers the electronic record, networks and security, reception and claims systems, the data used by managers and the connections to Japan's health-data systems. I share decisions with clinical leaders: they decide which services pause during an outage and what the record must capture, while I set out cost, risk and timing. The three cases on this site belong to this period. Each involved a colleague I respect disagreeing with me, and each left the group with a weakness it could see more clearly. I also hold the supplier relationships, which I review each year for any personal connection.

Reason for the move

Current appointment; exploring one outside directorship compatible with it.

Mandate

Bring three hospitals' separate systems, networks and support desks under one department.

Outcomes

• Joined the three hospitals on one network and a single service desk
• Connected the acute hospital to a prefectural consent-based record-sharing network
• Started quarterly reviews of returned claims with the medical affairs office before external audits

Decisions in context

Each hospital had its own systems staff, its own supplier contacts and its own idea of what support meant. My first year was spent listening to ward managers at the community and rehabilitation hospitals, who felt that the acute hospital always came first. We created one service desk with a published response standard and moved the three sites onto one network, which later made consolidation of the record possible. I also joined the prefecture's record-sharing network, in which participating hospitals and clinics can view a patient's shared information with consent. With the medical affairs office I began reviewing returned claims each quarter, which showed how often an error began at reception. The period taught me to manage people I could not see every day and to resist decisions that suited only the largest hospital. It also gave me the data that later persuaded the board.

Reason for the move

Appointed chief information officer and executive officer when the chairman created the post in 2020.

Mandate

Stabilise the hospital's electronic record and its support after a period of frequent outages.

Outcomes

• Reduced unplanned record outages from 14 in 2011 to 3 in 2014
• Wrote the hospital's first downtime procedures with the nursing department
• Renegotiated supplier maintenance terms to include an engineer stationed within two hours

Decisions in context

I joined weeks after the March 2011 earthquake, when the hospital was still reviewing how its systems had coped with power restrictions. The electronic record stopped without warning several times a year, and each ward improvised its own response. I began with the outage history and found that most failures traced to ageing servers and a maintenance contract that allowed engineers a full day to arrive. We replaced the equipment, renegotiated the contract and, with the director of nursing, wrote downtime procedures that every ward rehearsed once a year. This was the first time I worked directly for the people who used the systems rather than for the company that sold them. The difference changed how I measured success. A completed project mattered less than whether the night shift trusted the record and knew what to do when it disappeared.

Reason for the move

Asked to extend the same work across the group's three hospitals.

Mandate

Replace the chain's dispensing and claims systems without interrupting daily dispensing.

Outcomes

• Replaced the dispensing system across 118 pharmacies in phased weekend cut-overs
• Moved monthly insurance claims to online submission
• Reduced returned claims by recording eligibility problems at the counter

Decisions in context

A dispensing pharmacy cannot pause. Prescriptions arrive from nearby clinics and hospitals every day, and each must be dispensed, recorded and claimed correctly. I replaced the chain's ageing dispensing system pharmacy by pharmacy, over weekends, with a pharmacist from each store trained in advance. I also moved monthly claims to online submission. The returned claims taught me most: when a counter recorded the wrong insurer, the money arrived months later, if at all, and a pharmacist had to reconstruct the visit. We added a simple record of eligibility problems at the counter, which reduced returns. I learnt to design for part-time staff who rotated between stores and could not be trained for long. By 2011 I wanted to see where the prescriptions came from. A hospital, where the record and the decisions begin, was the obvious next step.

Reason for the move

Wanted to work inside a hospital, where the clinical record begins.

Mandate

Deliver electronic medical record implementations on time and within the agreed scope.

Outcomes

• Led implementations at nine hospitals
• Introduced ward walk-throughs with night-shift staff before each go-live
• Limited customisation in my last two projects to a priced list agreed with hospital committees

Decisions in context

Electronic records were new to most Japanese hospitals when I began this role, and each customer wanted the screens to reproduce its paper forms. We priced every change, and the customisations accumulated. By my fourth project I could see what this meant for the hospital five years later: an upgrade that required repeating every change, and a dependence on our engineers. In my last two implementations I persuaded the hospitals' committees to agree a short, priced list of changes in advance and to accept the standard design elsewhere. I also began walking the wards with night-shift nurses before go-live, which found problems that daytime testing missed. The work gave me a supplier's understanding of contracts, margins and support. It also left me uneasy about handing a system over and leaving, which is why I eventually crossed to the customer's side.

Reason for the move

Wanted to own the result after go-live rather than hand it over.

Mandate

Build and install order-entry interfaces between wards, pharmacy and laboratory.

Outcomes

• Built pharmacy and laboratory interfaces for five hospitals
• Supported ward go-lives through night shifts
• Documented ward workarounds for the product design team

Decisions in context

I joined straight from my engineering degree and spent five years writing and installing the software that carried orders from wards to the pharmacy and laboratory. Much of the work was done at night, during go-lives, when I stood beside nurses using what we had built. The three nights on a surgical ward in 1996 were typical: the system worked as specified, and the staff found faster ways around it. I began writing down those workarounds and sending them to our product designers. A few changed the product; most did not. I learnt the structure of hospital orders, the pressure on pharmacy staff and the gap between a test script and a ward at two in the morning. The role gave me technical depth and patience with detail. It also showed me that the people who use a system rarely write its specification.

Reason for the move

Promoted to project manager as hospitals began adopting electronic records.