The eating plan made in hospital often never makes it home. A review of 13 trials asks what actually helps older patients after discharge

An older person who leaves hospital underweight or eating too little goes home tired, often less mobile, and sometimes alone. In hospital someone noticed the problem: a screening score, a dietitian visit, maybe a supplement drink on the tray. The question in a mini review published on 22 October in Frontiers in Nutrition is simple. What happens to that care once the patient is through the front door?

The record is not encouraging. The authors, from two hospitals in Ningbo, China, gathered audits from routine care. In an Australian group of 42 older medical patients at nutritional risk, six (14%) were seen by a dietitian after discharge and eight (19%) got practical nutrition help at home. A second Australian audit of 345 admissions of malnourished older patients found nutrition recommendations in the electronic discharge summary for 35 of them (10%), and dietitian follow-up recorded within six months for 79 (23%). In a US chart review of 76 malnourished adults, five (6.6%) had written instructions to keep taking supplement drinks. The authors note that missing paperwork does not prove no advice was given, but the next doctor or nurse cannot act on advice they never see.

They then compared 13 randomised trials that tried to carry the care home. The picture is mixed in a consistent way: eating, weight or strength often improved, while readmissions, disability and deaths mostly did not move. In a Danish trial of 208 people aged 75 and over who lived alone, dietitian home visits cut 90-day readmissions (hazard ratio 0.4), but the visits lasted 45 minutes against 15 for the phone calls, whose result was inconclusive, and 88 of the 208 had a broken hip. Another Danish package of a plan, a starter food box, electronic messages to the local council and follow-up raised the share of people meeting at least three-quarters of their energy and protein needs at 16 weeks to 82%, against 61%, yet six-month readmissions were the same. A Swedish trial of 671 people found no survival difference over a median 4.1 years. In the US NOURISH trial of 652 patients, a high-protein supplement did not reduce the main combined result of death or readmission, although deaths at 90 days were lower (4.8% against 9.7%).

Part of the problem is plain follow-through. In a Scottish trial, reported adherence to the supplement drink was 38% (50% for the lower-calorie control drink); in another Danish study, the parts of the plan actually delivered ranged from none to 61%; in a French home-visit programme, 42% of 351 older adults completed every planned visit. From all this the authors sketch a six-step loop: check goals, food access and who can help at home; send a handover the next service can use; get that service to confirm it has taken the patient on; make early contact; agree who acts if eating slips; and audit the whole thing. They are clear that this is a proposal, not a tested pathway. It is also a focused narrative review rather than a systematic one, with no formal risk-of-bias grading, searches limited to PubMed and Europe PMC, and study selection the authors say was AI-assisted and not double-checked. For families, the practical point stands without the framework: ask before discharge who will check on eating at home, and when.