Access a health centre can staff every week

In 2022 Solvreten spent SEK 92 million on agency doctors, almost all of it at eleven health centres in our northern region. Several of those clinics had not employed a permanent GP for more than two years. Patients with diabetes or heart failure met a different doctor at nearly every visit, and our salaried doctors spent hours each week briefing colleagues on short assignments.

The stakes went beyond cost. The region expected providers to reduce agency use, and our majority owner questioned whether the northern clinics were viable at all. Ending agency cover too quickly would have left patients without timely assessment.

I proposed ending agency contracts at eight of the eleven clinics over fifteen months. We recruited fourteen salaried GPs with protected development time and supervision duties for trainee doctors, formed a six-doctor video team for structured follow-ups across the region, and widened nurse assessment under protocols the chief medical officer approved. The hardest part was reducing two satellite units from five opening days to three. Our northern operations director argued that patients would move to competitors, and both local heads of clinic opposed the change. The region's contract manager asked how access would be protected. We did give up listings: about 2,300 patients moved to other providers, costing roughly SEK 7 million a year in capitation.

By 2024 agency spending had fallen to SEK 38 million. The share of patients receiving a medical assessment within the time limit that applies across Sweden fell from 89 to 81 percent for six months, then recovered to 90 percent by the end of 2024. Visits with the patient's own named doctor rose from 41 to 55 percent across the eight clinics, and the north now has a stable medical team.

I would do two things differently. I would have gone to the region before, not during, the decision on satellite opening days. And I would have kept one satellite open four days rather than three, because the capitation lost there exceeded the staffing saved.

By 2023 all three of our regions were encouraging digital contact, and our platform supplier proposed moving every first contact into structured chat triage, with the telephone used only for callbacks requested through the app. The model would have released about twenty nurse posts from telephone work and saved roughly SEK 11 million a year.

I accepted most of the argument but not all of it. Our own figures showed that patients over seventy-five made about a third of our contacts and that most of them called. People with limited Swedish, and relatives phoning for a parent, also relied on the telephone. If the only door was an app, I expected many of those patients to turn up at the emergency department instead.

Our chief financial officer and head of digital argued for the full model, and the operating partner from our majority owner asked why we were paying for a channel patients were leaving. I proposed a mixed design: digital booking, renewals, test results and chat triage for everyone who wanted them, and nurse-staffed telephone triage in four regional contact hubs with same-day callback. The same nurses worked both queues, so capacity followed demand through the day. We gave up the full saving and accepted a slower reduction in contact-centre staffing.

By the end of 2025, 52 percent of contacts arrived digitally and 96 percent of telephone callers were rung back the same day. Among patients over seventy-five, 71 percent still chose the telephone. One finding surprised us: digital chats were followed by a further contact within a week in 18 percent of cases, against 9 percent after telephone triage. We rewrote the chat questions for three common complaints, and that rate has since fallen to 13 percent.

Next time I would measure repeat contact for each channel from the first week, and I would begin in four health centres instead of introducing the design across all forty-one at once.

In the summer of 2024, six health centres in our central region pooled prescription-renewal requests into one shared queue to cover holiday absence. By late July, 640 requests had waited more than five days. A patient with heart failure went nine days without a renewed diuretic and was admitted to hospital. Our chief medical officer investigated and reported the event to the Health and Social Care Inspectorate under the Patient Safety Act.

The investigation found that no one at the six clinics regarded the shared queue as theirs, and that two colleagues had noticed the backlog without reporting it as a deviation. Our reporting tool was slow, and many staff believed reports were used to judge them.

I took two decisions. First, we ended pooled summer queues and funded a permanent renewal team of four pharmacists and six nurses, costing SEK 5.6 million a year and removing about SEK 3 million of summer staffing savings. Second, with the chief medical officer, I set out to increase internal reporting: a two-minute form, feedback to every reporter, and no use of reports in individual performance reviews. The board's quality committee initially wanted a goal to reduce the number of reports. I argued against it and proposed measuring time to documented action and the severity mix instead. One region's contract monitor later asked why our reports had more than doubled.

Reports rose from 1,240 in 2023 to 2,910 in 2025. The share closed with a documented action within sixty days rose from 48 to 79 percent, and the median renewal turnaround fell to 1.8 days. A later inspection of two of our clinics found the reporting system working but criticised how one clinic documented follow-up of its investigations. We accepted the finding and changed that routine.

I would do one thing differently. I would have explained the reporting change to all three regions before the numbers rose, rather than answering a contract monitor's question after the figures had already moved.