The care planning cycle explained: assess, plan, implement, evaluate

The care planning cycle is the loop every care plan is supposed to run on: assess, plan, implement, evaluate, and back to assess. It is taught on every care course in the country and it is the single most useful idea in care planning, because it makes clear that a plan is not a document you write on admission and file. It is a claim about what will help someone, tested every day on shift and checked every month against what actually happened. This guide takes each stage in turn, sets out what belongs in the record at each point, gives timescales, works through two real examples, and explains where the cycle breaks in most homes and how to fix it.

The short answer

Four stages. Assess: find out what the person needs, wants and is at risk of, using conversation, observation, records from other services and validated tools. Plan: write down what will be done, by whom, how often and what it is meant to achieve, in the person's own terms. Implement: deliver it on every shift and record what happened in the daily notes and charts. Evaluate: compare what happened with what the plan intended, decide whether it is working, and change the plan if it is not. Then reassess and go round again. Formal reviews happen monthly for the evaluation and at least annually in full, with an immediate reassessment after any significant change. A home where the cycle works has care plans that look different this month from last month. A home where it does not has beautiful plans that nobody has changed since admission.

Where the cycle comes from

It began as the nursing process in the 1970s, taught as APIE: assess, plan, implement, evaluate. Some courses teach ASPIRE, which adds a systematic assessment of needs and a recheck stage: assess, systematic nursing diagnosis, plan, implement, recheck, evaluate. Social care adapted it into person-centred planning, and the Care Act 2014 wrote a version of it into law for local authorities: assess needs under section 9, decide eligibility under section 13, produce a care and support plan under sections 24 and 25, and keep it under review under section 27.

The names differ, the logic does not. Find out, decide, do, check. The value is in the arrow that goes back to the start.

The stages at a glance

Stage one: assess

Assessment starts before the person arrives. A pre-admission assessment done properly answers whether you can meet the person's needs at all, which is a regulatory question as well as a practical one. Read the hospital discharge summary, the local authority care and support plan, the previous provider's records, the medicines list and any specialist reports. Then go and meet the person.

The records tell you about conditions. The person tells you about life. The relative tells you about the twenty years the records do not cover. You need all three, and the assessment should say where each piece of information came from and when.

Within the first 24 hours you need enough to keep someone safe: mobility and transfers, skin, continence, nutrition and swallowing, medicines, communication, cognition, mental health, risk of falls, what they like to eat, how they like to be addressed, who matters to them, and what a bad day looks like. The rest can follow in the first week.

What a good assessment gathers

• The person's own account of what they need help with and what they want to keep doing themselves
• Life history: work, family, faith, culture, routines, what they are proud of
• Communication: how they express pain, distress, yes and no
• Health conditions, current treatment and who is involved from outside
• Medicines, including what they take, how they like to take them, and any covert or refusal history
• Validated risk tools, with scores and dates
• Capacity for the decisions that matter now, and any attorney, deputy, advance decision or advance statement
• What has gone wrong before, and what helped
• What the person wants life here to look like in three months

The last one is the question homes ask least and inspectors ask most.

Assessment tools that feed the plan

Tools are not the assessment. They are structured questions that produce a score, and a score is only useful if it changes something. Use MUST for nutrition, Waterlow or a similar tool for skin, a falls risk assessment, a moving and handling assessment, a continence assessment, an oral health assessment, and a capacity assessment for specific decisions.

The test of whether you are using them properly is simple: open the plan and see whether the score appears as an action. A MUST score of 2 with no food chart, no weekly weights, no fortified diet and no GP referral is a number in a box. Details on calculating and acting on the score are in how to calculate a MUST score, and the wider risk record is covered in risk assessments in social care.

Stage two: plan

A care plan translates assessment into instructions that a worker on their first shift could follow, and into outcomes that let you tell later whether it worked. Every plan should say what the person can do themselves, what they need help with, exactly how that help is given, how often, who by, and what result you are aiming for.

The structure I use for each need is five lines: what the need is, in the person's words where possible; what we are trying to achieve; what staff will do, specifically; what we will watch for; and when this will be evaluated. Five lines per need, ten to fifteen needs, and the plan is readable in ten minutes, which matters because a plan nobody reads changes nothing.

Write it with the person, not about them. Regulation 9 on person-centred care requires that care and treatment is designed with the person's involvement and reflects their preferences. A plan that could be about anyone with the same diagnosis has failed that test. There is a full walkthrough in how to write a person-centred care plan.

Outcomes, not tasks

The difference between a task and an outcome is the difference between a plan that ages well and one that does not. "Assist with personal care each morning" is a task. It is true on day one and day one thousand and it tells you nothing about whether anything is working.

"Jordan washes and dresses himself with prompting, and wants to be able to shave without help by the summer" is an outcome. It says what good looks like, it can be evaluated, and it can fail, which is what makes it useful. Keep tasks in the plan, because staff need to know what to do, but hang every group of tasks under an outcome that someone can be held to.

Outcomes should be the person's, not the service's. "Reduce falls" is your outcome. "Get to the day centre on Tuesdays without being frightened of falling" is theirs, and the second one produces better care.

Stage three: implement

This is the stage people talk about least and where the cycle most often fails. A plan becomes practice through four routes: handover, the task list or daily schedule, supervision, and the worker actually reading the plan.

Make the plan visible where the work happens. If a worker has to walk to the office, find a folder and read fourteen pages to learn that Priya takes her tea black and needs her left hearing aid in first, they will not do it. A one-page profile at the front, or the equivalent on a handheld device, is what gets read.

Handover is where the plan meets today. A good handover says what changed, not what is always true, and names who is doing what this shift. Recording is the other half: the daily note is the evidence that the plan was implemented, which is why vague notes destroy the cycle. "Personal care given, all care needs met" proves nothing. The article on daily care notes examples shows what a usable note looks like.

The gap between the plan and the shift

Every home has one. The plan says two-hourly repositioning and the chart shows four-hourly at night. The plan says encourage Marcus to help with cooking and nobody has time. The plan says a walk in the garden every day and it rained all week.

The gap is information, not failure. If the plan cannot be delivered with the staff you have, either the plan is wrong or the staffing is wrong, and both are things a manager needs to know. The wrong response is to leave the plan as an aspiration and let staff quietly not do it, because that teaches the team that the plan is fiction. Ask at every evaluation: was this actually delivered, and if not, why not?

Stage four: evaluate

Evaluation is not a summary of the month. It is a judgement against each plan: is this working, what is the evidence, and what changes.

A usable evaluation answers five questions for each need. Was the plan delivered as written? What did the records show: the charts, the weights, the incidents, the MAR, the daily notes? Has the outcome been achieved, partly achieved or not achieved? What does the person say about it? What changes now, or is the plan confirmed as it stands?

Written properly that is a paragraph, not a page. "Repositioning delivered as planned on 27 of 30 days, with three gaps on nights of 4 to 6 March when the unit was short. Sacrum remains intact, Waterlow unchanged at 17. Priya says the new cushion is more comfortable. Plan unchanged, night gaps raised with the deputy and covered from 10 March." That paragraph proves delivery, quotes evidence, includes the person, and names an action.

The phrase to ban is "no change". It appears in thousands of care files every month and means nobody looked.

Stage five: review and reassess

A review is a bigger event than an evaluation. It brings the person, the family or representative where the person wants them, and the professionals involved, and it asks whether the whole plan still fits.

Full reviews happen at least annually, and in practice most homes review at four to six weeks after admission and then annually with monthly evaluations in between. What matters more than the calendar is the trigger list: any fall with injury, any hospital admission or discharge, a new diagnosis, a medicines change, weight loss or gain, a new pressure area, a change in mood or behaviour, a safeguarding concern, a change in capacity, a change in what the person wants, or a request from the person or their family. Any of those means reassess now, not at the next scheduled date. How often care plans should be reviewed covers the timing question in detail.

Timescales that work in a care home

A worked example: Jordan

Jordan is 27, has a learning disability and anxiety, and has just moved into a four-bed supported living service. He wants to go to college and he has stopped using buses after a frightening incident last year.

Assess. The team learns that Jordan can travel independently when a route is familiar, that his anxiety peaks with crowds and unpredictable delays, and that he has an activity he loves, watching trains. His own words: "I want to go to college but I can't do the bus."

Plan. The outcome: Jordan travels to college independently by the end of the autumn term. Interventions: a staff member travels with him at the same time each Tuesday and Thursday for four weeks; the route is written as a picture guide; a scripted plan for what to do if the bus is late; a weekly conversation about how it felt, scored out of ten. Watch for: avoidance, sleep, refusal to leave the flat. Evaluate: monthly, with a check at two weeks.

Implement. The travel support goes into the rota so it is a named person's job, not an aspiration. Staff record each journey in one line: who went, where they sat, how Jordan rated it, what happened.

Evaluate. Month one: eight journeys completed, ratings moving from 3 to 6, one journey abandoned when the bus was diverted. Jordan says the picture guide helps and the scripted delay plan does not, because he cannot read it when he is panicking. Change: the delay plan becomes a card in his wallet with three short lines and a phone number. Month two: staff member sits two rows behind rather than beside him. Month three: Jordan travels alone with a phone check-in at each end.

That is the cycle working. The plan was wrong in one specific way, evaluation found it, and the fix was small.

A worked example: Priya

Priya is 88 and has vascular dementia. Over six weeks she loses 4.3kg.

Assess. MUST score rises to 2. Observation shows she leaves most of her lunch and eats well at breakfast. Her daughter says she has always disliked eating with strangers. The dentist finds a sore under her denture.

Plan. Outcome: Priya maintains or regains weight, aiming for 48kg by the end of June. Interventions: dentist first, soft options while the sore heals, lunch served in her room or with the two residents she knows, fortified milk twice daily, finger foods available all afternoon, weekly weights on Mondays, food chart for four weeks. Watch for: refusals, swallowing difficulty, coughing during meals. Evaluate: fortnightly while the food chart runs.

Implement. The kitchen gets the fortification instruction in writing, the room change goes on the handover sheet and the seating plan, the weights go into the Monday routine with a named person.

Evaluate. Fortnight one: weight stable, food chart shows lunch intake improved from a quarter to three quarters, denture sore healing. Fortnight two: weight up 0.6kg. Month two: weight up 1.4kg overall, MUST back to 1, food chart stopped, weekly weights continue for another month then monthly. Plan updated to keep the seating arrangement permanently, because that turned out to be the thing that mattered.

Nobody needed a dietitian in the end. They needed to find out why she was not eating, which is stage one.

Care and support plan, care plan, support plan: which is which?

These get used interchangeably and they are different documents with different authors.

• Care and support plan. Written by the local authority under sections 24 and 25 of the Care Act after a section 9 assessment and a section 13 eligibility decision. It records eligible needs, the outcomes the person wants, the personal budget and how needs will be met. The council reviews it under section 27.
• Care plan or support plan. Written by you, the provider, and far more detailed. It says how care is delivered day to day. This is the one CQC inspects.
• Health care plans. Written by clinicians: wound care plans, diabetes plans, epilepsy protocols, end of life plans such as ReSPECT.

Your plan should reference the council's outcomes, not ignore them, and a mismatch between the two is worth raising with the social worker. If the council's plan funds four calls a day and the person now needs six, the plan needs reviewing, not absorbing quietly.

Capacity within the cycle

Capacity is decision-specific and time-specific, so it belongs at every stage rather than as a form completed on admission. Assess capacity for the decisions the plan actually requires: consent to personal care, consent to medicines, use of bedrails, going out alone, sharing information with family. Where capacity is lacking, record the best interests decision, who was consulted, and the less restrictive options considered.

Revisit it at evaluation. Capacity fluctuates, particularly with delirium, infection and pain, and a plan built on an assessment from eleven months ago will not stand up.

Risk within the cycle

Risk assessments are part of the same loop. Assess the risk, plan the controls, implement them, evaluate whether they worked, and change them. What CQC now expects, under the quality statement on involving people to manage risks, is that the person is part of that conversation and that the response is proportionate.

A plan that removes all risk usually removes the life with it. If Jordan is never allowed on a bus alone, he is safe and he is stuck. The record should show the risk, the person's wishes, the discussion, the controls, and who agreed. That is positive risk-taking, and done properly it is defensible even when something goes wrong.

Where the cycle breaks

• Assessment done once. The admission assessment is never updated, so plans are built on a version of the person from two years ago.
• Plans copied between residents. Identical wording in three files tells an inspector everything.
• Tasks without outcomes. Nothing can be evaluated, so evaluation becomes "no change".
• Implementation invisible. Daily notes too vague to prove the plan was delivered.
• Evaluation as a signature. A date and initials with no judgement.
• Review by calendar only. A fall in week two waits until the annual review.
• The person absent. The plan is written about them, and their views appear nowhere.
• Charts that contradict the plan. The plan says four-hourly checks and the chart shows two entries a night.

Auditing the cycle

Audit the loop, not the paperwork. Pick three residents a month at random and follow one need each the whole way round: find it in the assessment, find the plan for it, find the evidence it was delivered in the charts and notes for the last two weeks, find the evaluation, and see whether anything changed as a result.

That audit takes about forty minutes for three residents and it tells you far more than a checklist asking whether every plan is signed. Record what you found and what you are changing, because the action is the part that counts.

How the cycle looks to an inspector

An inspector will pick a resident and follow exactly the same trail. They will ask a care worker what Priya needs at mealtimes and compare the answer to the plan. They will look at whether the weight loss in February produced a change in March. They will ask what the person wanted and where that is recorded. They will ask you when this plan last changed and why.

The strongest evidence is a plan that visibly moved: an outcome that was not achieved, an honest evaluation saying so, and a different approach tried. Plans that have never changed read as care that has never been thought about. This is the substance behind the quality statements on assessing needs, person-centred care and monitoring outcomes.

Paper and digital in the cycle

The cycle works on paper. It is just harder, because the evidence lives in four places: the plan in a folder, the charts on a clipboard, the notes in a daybook and the incidents in the office. Evaluating means collecting all of it first, which is why monthly evaluations get written from memory at ten at night.

A digital record helps at exactly one point: it puts delivery evidence next to the plan, so the evaluation can be written while looking at the charts, the notes and the incidents together. Systems like Kiwi link the plan, the daily record and the review in one place through care records, and they can flag when an evaluation is overdue. What no system does is decide whether the plan is working. That judgement is the job, and it is the part inspectors look at hardest.

Teaching the cycle to your team

Care workers meet the cycle in the Care Certificate and forget it, because it is taught as a diagram. Teach it as a story about one resident instead. Take a real anonymised plan, show the assessment, show the plan, show the notes for a week, show the evaluation, and ask the team what they would change. Then ask each worker to write one evaluation paragraph and read them out.

Key workers should own the loop for their residents: they know the person, they write the evaluation, and they raise the change. A key worker system with named residents does more for care planning quality than any template, and it gives you a person to ask when an inspector picks that file. Examples of what a strong plan looks like in a learning disability service are in person-centred care plan examples.

Common mistakes

• Treating the plan as a compliance document. It is a set of instructions for today's shift.
• Writing the plan before knowing the person. Day one plans should be short and honest about what you do not yet know.
• Evaluations that say nothing. "Care plan reviewed, no change required" is the most common sentence in UK care files and the least useful.
• Not recording what was not delivered. The gaps are the most valuable information you have.
• Excluding the person. Regulation 9 requires involvement, and it is the first thing an inspector tests.
• Too many plans. Thirty separate plans for one person will not be read. Ten to fifteen well-written ones will.
• Risk assessments on a separate cycle. They should be evaluated with the plan they support.
• No trigger list. Reviews happen by calendar and miss the events that mattered.

Final conclusion

The care planning cycle is four honest questions asked in order and then asked again. What does this person need and want? What are we going to do about it and what for? Is it being done? Is it working? Everything else is detail. Assess before admission and keep assessing. Write plans with outcomes that can fail, not just tasks that cannot. Make the plan visible where the work happens and prove delivery in the daily record. Evaluate monthly with evidence, name what did not work, and change something. Reassess immediately when anything significant changes rather than waiting for the calendar. Do that consistently and your care plans will tell the story of someone's life improving, which is the only evidence that really matters. If you want to see the cycle running in one record, book a demo.

Frequently asked

What are the stages of the care planning cycle?

Four stages: assess, plan, implement and evaluate, then back to assessment. Some courses teach five or six stages by splitting assessment into gathering information and identifying needs, or by adding a recheck step before evaluation. The logic is the same: find out, decide, do, check, and start again.

What does assess plan implement evaluate mean in care?

Assess means finding out what the person needs, wants and is at risk of. Plan means writing down what will be done, by whom, how often and what it should achieve. Implement means delivering it on every shift and recording what happened. Evaluate means comparing what happened with what was intended and changing the plan if it is not working.

How often should each stage of the cycle happen?

Assessment before admission, updated within 24 hours and again at four to six weeks. An interim plan on day one and a full plan within five to seven days. Implementation and recording are continuous. A written evaluation monthly. A full review at least annually, and an immediate reassessment after any fall with injury, hospital admission, new diagnosis, medicines change, weight change or change in what the person wants.

What is the difference between an evaluation and a review?

An evaluation is a monthly judgement by the key worker or named nurse against each care plan, asking whether it was delivered, what the records show and whether the outcome was achieved. A review is a larger event involving the person, family and professionals that asks whether the whole plan still fits, and normally happens annually or after a significant change.

What is the difference between a care and support plan and a care plan?

A care and support plan is written by the local authority under sections 24 and 25 of the Care Act 2014 after a needs assessment and eligibility decision, and records eligible needs, outcomes and the personal budget. A care plan or support plan is written by the provider and is far more detailed about how care is delivered day to day. It is the provider's plan that CQC inspects.

What makes a good care plan outcome?

It describes what good looks like for that person in their words, it can be evaluated, and it can fail. Get to the day centre on Tuesdays without being frightened of falling is an outcome. Assist with mobility as required is a task. Keep the tasks, because staff need to know what to do, but group them under an outcome someone can be held to.

Why do most care planning cycles break down?

Because evaluation becomes a signature. A plan is written well on admission, delivered reasonably, and then evaluated every month with the words no change required, so nothing is ever tested or altered. The second most common break is daily recording too vague to prove the plan was delivered at all.

Where does risk assessment sit in the care planning cycle?

Inside it, not beside it. Assess the risk, plan proportionate controls with the person, implement them, evaluate whether they worked, and change them. Risk assessments should be evaluated at the same time as the care plan they support, and the record should show the person's own wishes and the discussion about them, not just the control measures.

Sources

• Care Act 2014, sections 9, 13, 24, 25 and 27
• GOV.UK: Care and support statutory guidance, chapters 6 and 10
• Health and Social Care Act 2008 (Regulated Activities) Regulations 2014, Regulation 9: Person-centred care
• Care Quality Commission: single assessment framework and quality statements
• NICE guideline NG22: Older people with social care needs and multiple long-term conditions
• NICE guideline NG108: Decision-making and mental capacity
• BAPEN: Malnutrition Universal Screening Tool (MUST)
• Skills for Care: Care Certificate and person-centred care standards