By the time Malik returned to Mwangaza, the anonymous complaint had already been printed and pinned to the ward noticeboard. It was only three lines long: `Doctors are afraid to refer because administration calls too many referrals a failure. Kevin is not the first. People will be afraid to speak.` Beneath it, someone had added in pen: `And who is responsible when you refer without good reason?`
People passed the board pretending not to read it. Malik knew everyone already had.
Achieng found him there. “I tried to take it down before the media saw it.”
“Don’t.”
“Why?”
“Because it already exists. If we hide it now, we prove its point before we even know whether the claim is true.”
Four hours later, the debrief began in the ward conference room. Wekesa sat at the far end of the table with her arms folded. Musa, Achieng, two junior doctors, and a clinical officer were present. Nobody wanted to speak first.
Malik wrote three rules on the board: no unnecessary patient names; no punishment for anyone describing workflow; every claim must be separated from evidence.
Kamau, one of the doctors, asked, “Is this an investigation?”
“A workflow debrief.”
“What’s the difference if the report ends up with the board?”
“The difference is that we haven’t started with a conclusion.”
Kamau gave a dry laugh. “That sounds like the language of people who write reports.”
Wekesa leaned forward. “And the distinction matters. This complaint implies that management blocks referrals. That is not true.”
Musa spoke quietly. “Nobody is saying management phones us and says, ‘Do not refer.’”
“Then what is the problem?” Wekesa asked.
Musa looked at Malik, then at Achieng. “It’s what happens afterward. If you refer, people ask why you couldn’t manage the case. If you keep the patient and things go well, nobody asks anything. The pressure points in one direction.”
The room went still.
Kamau added, “Last month I referred a child with breathing problems. The review said the child could have stayed here. I got three calls about an ‘unnecessary transfer.’ The next week I had another case and kept the patient longer. No harm happened, but I knew I was thinking about the review instead of only the patient.”
Wekesa frowned. “Why didn’t you say that?”
Kamau met her eyes. “Because the review was the way of saying it.”
Malik wrote `feedback asymmetry` on the board.
Achieng said, “Nurses have the same problem. A doctor says ‘observe,’ but nobody tells us when a nurse is allowed to say observation has reached its limit. Call a consultant early and you look like you have jumped the hierarchy.”
“Are there criteria?” Malik asked.
“For some conditions, yes. But not for ownership of referral. That part is culture.”
Nyambura, another clinical officer, said, “And the red phone itself is a bottleneck. One person can be busy on another call. If Bahari doesn’t answer, nobody owns the next step. Everyone assumes someone else will try again.”
Wekesa pointed at the complaint. “None of this proves the KPI wording discourages referrals.”
“It doesn’t,” Malik said. “But it shows an environment where ambiguous wording can have an effect.”
“What do you want me to do with that sentence?”
“Don’t defend it just because you wrote it. Test it.”
Wekesa held his gaze. “And you don’t build a case against it just because you’ve decided it is a problem.”
“Agreed.”
Malik erased `KPI` and replaced it with a question: `What makes staff delay escalation?`
Answers began appearing: fear of looking inexperienced; uncertainty about who makes the final call; bed shortages at Bahari; ambulance coordination; family trust; documentation burden. Then someone added an unsigned sticky note: `fear of being called the doctor who refers too much`.
Wekesa read it without comment.
Malik refused to let the session become pure venting. He asked for concrete actions, not feelings alone. Musa described the missed callback. Achieng described a case where a nurse saw deterioration but waited for a doctor in theatre because she did not know whether she could contact a referral consultant directly. Kamau described a review that counted transfers without showing the severity mix of the cases.
“So where is our loyalty?” Wekesa asked suddenly. “If this hospital is seen as incompetent, funding falls, staff leave, and the same patients suffer.”
Achieng answered before Malik could. “Protecting the hospital does not mean hiding the part that fails us.”
“And exposing every weakness publicly is not quality improvement either,” Wekesa said.
“Nobody said everything has to go on the radio,” Achieng replied. “But inside these walls, we have to be able to say it.”
Malik saw that the divide was not between good people and bad people. Every side was protecting something legitimate—safety, reputation, resources, trust—and those things were colliding.
He proposed three days of anonymous workflow submissions built around four questions only: when was referral initiated, who owned the next action, what delayed it, and what pressure did the staff member feel? No patient identifiers. No blame fields.
“You’ll get a hundred complaints,” Kamau said.
“We’ll analyze patterns, not mood.”
“Who sees them?” Wekesa asked.
“Farah from quality, me, and one nursing representative. Raw submissions do not leave the review group without a reason.”
After a long argument, Wekesa agreed to a three-day pilot.
When the session ended, people left in small groups. Kamau did not greet Malik. Musa gave him a small nod. Achieng stayed behind to help wipe the board.
“You understand now?” she asked.
“Understand what?”
“People will think you have chosen a side.”
“And you?”
“I think you still haven’t chosen one. That may make everyone angry.”
Malik laughed softly, picked up his laptop, and walked toward his office. The corridor was quiet. Under his door lay a sheet of paper folded twice.
He opened it.
There was no name, signature, or explanation.
`Did you come back to treat us, or to open a case against us?`
---