Dysphagia and IDDSI in learning disability homes: levels, thickeners and what to record

Dysphagia is swallowing difficulty, and in learning disability services it is not an occasional complication. It is one of the most common and most lethal health risks the people we support live with, and many of them cannot tell you that swallowing hurts or feels frightening. The job of the service is to get the right texture in front of the person at every meal, recorded in a way that cannot be lost.

The short answer

Everyone with a swallowing risk needs a named IDDSI level for food and a named IDDSI level for drinks, set by a speech and language therapist, written on the front of the care record, checked in the kitchen with the IDDSI test, and recorded at every meal. Those are four separate jobs and most services do two of them. The level is set, it is written somewhere, and then the kitchen guesses and the recording does not say what was actually served. This article covers the IDDSI framework in full, how levels get lost, what a good SALT recommendation looks like, the thickener errors that cause harm, what to watch during a mealtime, and exactly what to record so that a bank worker on a Sunday evening cannot get it wrong.

What is dysphagia and why is it so common in learning disability services?

Dysphagia is difficulty with any stage of swallowing, from getting food into the mouth to moving it safely past the airway. Research puts the rate at around one in three adults with a learning disability, rising above eight in ten in people with profound and multiple learning disabilities.

The reasons stack up. Cerebral palsy affects the muscles of the mouth and throat. Reflux is extremely common and often untreated. Long-term antipsychotics such as risperidone, olanzapine and haloperidol cause sedation, dry mouth and sometimes drug-induced parkinsonism, all of which slow the swallow, and anticonvulsants add more sedation. Poor dentition means food is not chewed. Postural problems mean people eat slumped. Some people eat very fast, cram, or take food from other plates. Any one of these turns a meal into a risk. Several together, which is the usual picture, make a mealtime a clinical procedure that happens to be social.

Why choking deaths in learning disability services keep happening

Choking is a leading cause of avoidable death in learning disability services, and the reviews that follow nearly always find the same thing: the right recommendation existed and did not reach the person who served the food. LeDeR reviews, coroners' Prevention of Future Deaths reports and safeguarding reviews repeat it year after year.

The recommendation is in a letter in a folder. The agency worker has not read the folder. The level changed three months ago and the kitchen whiteboard still shows the old one. The person went to a day service and was given a sandwich. Nobody was careless in an obvious way. The system simply had too many places for a single number to hide.

The IDDSI framework explained

IDDSI is the International Dysphagia Diet Standardisation Initiative, a global framework of eight levels numbered 0 to 7 that describes how thick a drink is and how soft and how small a food is. It replaced the old UK descriptors such as Stage 1, Stage 2 and fork mashable, which meant different things in different places.

Levels 0 to 4 cover drinks and levels 3 to 7 cover foods. Levels 3 and 4 appear in both, because a moderately thick drink and a liquidised food are physically the same thing. The correct way to write a level is number and name together: Level 5 Minced and Moist, not just Level 5 and not just minced.

What makes IDDSI useful is not the naming. It is that every level has a physical test anyone can do in a kitchen in under a minute with a syringe, a fork and a spoon. A level without a test is an opinion.

The full IDDSI 0 to 7 table

How do you test an IDDSI level in the kitchen?

You test drinks with a 10ml slip tip syringe and a stopwatch, and foods with a fork and a spoon. The flow test takes ten seconds, the fork drip test and spoon tilt test take about the same, and any support worker can be taught all three in a single session.

For the flow test, block the nozzle of a 10ml slip tip syringe, fill it to the 10ml line, release for exactly ten seconds, then read what is left. The remaining volume gives you the level. For Level 4 and pureed food the syringe is no use, so you use the spoon tilt test: a full spoonful should hold its shape, then slide off cleanly in one lump when the spoon is tilted, leaving only a smear. If it runs off like a sauce it is Level 3, not Level 4.

For Levels 5 and 6 you use the fork. The gap between the prongs of a standard metal fork is about 15mm, which is exactly the maximum piece size for Level 6 in adults. Level 5 lumps must be no more than 4mm, about the width of one prong. Press with the side of the fork until the thumbnail blanches: if the piece does not squash, it is not compliant. Keep a labelled fork and a pack of syringes in the kitchen, not in a cupboard in the office.

Who sets the level, and who can change it

Only a speech and language therapist sets or changes an IDDSI level. Nobody in the service, including the registered manager, the nurse, the chef or the family, can move a person up or down a level, and a hospital discharge letter does not override a community SALT recommendation without a new assessment.

This needs writing down as a rule, because the pressure to drift is constant and it is always well meant. The person hates pureed food. They cried at lunch. They managed a biscuit at their sister's house and nothing happened. None of those is a clinical assessment. What the service can and must do is report the observation to the SALT, ask for a review, and record that it did so.

Downgrading matters as much as upgrading. A chest infection, a seizure cluster, a new sedating medicine or a spell of poor alertness can all make a swallow deteriorate quickly, and a chest infection should trigger a same-week SALT referral, because aspiration pneumonia is often the first sign a level is wrong.

What a SALT recommendation should actually say

A usable SALT recommendation names the food level, the drink level, the thickener product and the exact scoops per volume, the posture, the supervision level, the pacing, the cup or beaker to use, the foods to avoid and the review date. Anything less leaves the service guessing.

When a letter arrives that just says soft diet and thickened fluids, send it back and ask these questions:

• What is the IDDSI number and name for food, and for drinks, written separately?
• Which thickener, and how many scoops to how many millilitres, for which drink types? Hot drinks, fizzy drinks and fruit juice do not all thicken the same way.
• Does the person need full supervision, line of sight, or none?
• What posture: upright at 90 degrees, feet supported, and for how long after the meal?
• Which utensils and cup? Many people should never use a straw and many should never use a spouted beaker, because both bypass the person's control of the flow.
• Which specific foods are banned? Name the actual foods, not categories.
• What signs would mean stopping the meal and seeking advice?
• When is the review, and what would trigger an earlier one?

Thickener errors and how they happen

Thickener errors are among the most common medicines-adjacent incidents in learning disability services, and almost all of them come from three causes: the wrong scoop count, a switched product, and a drink left to stand.

The scoop is not universal. A scoop of one product does not thicken the same volume as a scoop of another, and the scoops themselves differ in size. If a service switches brand because of a pharmacy supply problem and keeps the old scoop count, every drink that person gets is at the wrong level until someone notices. Record the product name and the scoops per 200ml on the profile, and treat a product change as a change requiring the SALT and the pharmacist to confirm the new measure.

Starch-based thickeners such as the original Thick and Easy and Nutilis Powder continue to thicken as they stand, and can be broken down by amylase in saliva if a person back-washes into the cup. Gum-based products such as Nutilis Clear and Resource ThickenUp Clear are more stable and stay clearer, which matters for people who refuse cloudy drinks. Whichever is prescribed, make the drink, wait the stated time, check the level and remake it if it has been standing, because a jug of thickened squash made at breakfast and offered at eleven is not the level on the plan. Measure with the product's own scoop, levelled, into a measured volume. A mug is not 200ml.

Where the IDDSI level must appear so it is never missed

The level must appear in at least five places: the front summary of the person's record, the mealtime section of the care plan, the kitchen, the hospital passport, and anywhere the person eats away from home. If it lives in only one of those, the day it is needed will be the day the record is not there.

In practice that means a flag on the person's profile that shows before anything else opens, so a bank worker who has never met Jordan sees Level 5 Minced and Moist, Level 2 Mildly Thick, full supervision, first. It means a printed sheet in the kitchen with the person's photograph, the levels, the thickener recipe and the banned foods. It means the hospital passport carrying the same numbers, because wards otherwise apply their own assumptions. And it means the day service, the college, the respite unit and the family all having the same dated sheet.

This is where a learning disability care management platform earns its keep over paper, because the level lives on the profile once and appears on every screen, every handover and every printed summary at the same time. A system where the IDDSI level is typed into a free-text note is no better than a folder.

Jordan: a worked example of a level that went missing

Jordan is 29, has cerebral palsy and a moderate learning disability, and lives in a four-person supported living house. His recommendation is Level 5 Minced and Moist food, Level 2 Mildly Thick drinks made with two levelled scoops of Nutilis Clear per 200ml, upright at the table with feet flat, one teaspoon at a time, no straws, and no bread, no rice, no skins and no mixed consistencies such as soup with pieces in it.

In August his level changed from Level 6 to Level 5 after a chest infection. The SALT emailed the manager, who saved the letter and told the team at the Monday meeting. Two weeks later Jordan was given a Level 6 chicken casserole by a bank worker who had read the kitchen sheet, which still said Level 6. He coughed, went grey and needed back blows. The record said choking incident, resolved, and nothing else.

Three things changed. The kitchen sheet became a printout from the record rather than a card typed by hand, so it could not go out of date. The level change became a task with a named owner and a completion tick, not an item in the minutes. And the incident form gained a field asking what texture was served against what texture was prescribed, because without it nobody could see that the near miss was a texture error rather than bad luck.

Mealtime observation: what to watch for

A support worker at a mealtime is carrying out a clinical observation, and they need to know the specific signs: coughing, throat clearing, a wet or gurgly voice after swallowing, watering eyes, a change in colour, food pocketed in the cheek, multiple swallows for one mouthful, breathlessness, and a refusal to continue.

The wet voice is the one most often missed. Ask the person to say something after a few mouthfuls: a voice that sounds bubbly suggests material sitting on the vocal cords. Watering eyes and a red face during a meal are usually the airway reacting, not emotion. Pocketing food in the cheek means the person is not clearing, and the pocket can be inhaled twenty minutes later, which is why a mouth check at the end of a meal belongs in some people's plans.

What is silent aspiration and how would you spot it?

Silent aspiration is when food or fluid enters the airway without triggering a cough, so the person shows no obvious sign at the table. It is common in people with learning disabilities and it is dangerous precisely because the mealtime looks fine.

You spot it indirectly, over weeks, from the pattern rather than the meal: repeated chest infections, unexplained low-grade temperatures, a wet chest in the mornings, gradual weight loss, increasing tiredness, refusing food that was previously accepted, and a rising number of courses of antibiotics. Any of those should prompt a SALT referral and a GP review, and the service should be able to produce the evidence rather than describe an impression. That is the argument for recording every meal rather than only the difficult ones.

Posture, pacing and the environment

Posture is part of the prescription, not a preference. Most people need to be upright at about 90 degrees with feet supported, head in midline and slightly chin down, and should stay upright for 20 to 30 minutes after eating to reduce reflux and the risk of aspirating a returned mouthful.

Pacing is the other half, and rushing usually comes from the staffing pattern rather than the person. If one worker is supporting three people to eat, the meal will be paced to the worker. Put the time required in the plan: if Marcus needs 40 minutes and one-to-one support at lunch, that is a staffing fact for the rota and the commissioner's hours, not an aspiration. The environment matters too: television off, one conversation at a time, and the worker sitting beside the person at eye level rather than standing over them. For autistic people a predictable place and order to the meal reduces the anxiety that makes cramming worse, and our guide to autism friendly care planning covers the sensory side.

Medicines and dysphagia

A person on a modified texture diet cannot usually take ordinary tablets safely, and the answer is a pharmacist review of every medicine, not a decision at the trolley to crush things. Crushing a tablet is almost always outside the terms of its licence and can change how the drug works.

Modified-release and enteric-coated products, including modified-release metformin and modified-release levodopa preparations, release their whole dose at once if crushed. Epilepsy medicines matter especially here because so many people take them: levetiracetam, sodium valproate, carbamazepine and lamotrigine all have liquid or dispersible forms, and a switch between formulations can change the blood level, so it needs the prescriber, not the home. Thickened fluids interact too, because giving a tablet in thickened fluid can slow dissolution. Ask the pharmacist to review the whole list against the IDDSI level and record the agreed method for each medicine. Our article on PRN medication protocols covers recording as-required medicines properly.

Priya: crushing, capacity and the covert question

Priya is 46, has Down's syndrome and early onset dementia, and is on Level 4 Extremely Thick food with Level 3 Moderately Thick drinks. She takes donepezil, levothyroxine and a laxative, and she has begun refusing tablets, turning her head away and spitting them out.

The wrong response is to hide them in yoghurt. That is covert administration, it requires a capacity assessment and a best interests decision involving the GP and the pharmacist, and it must be recorded and reviewed. The right sequence is: assess capacity for this specific decision, involve Priya's sister and her advocate, ask the pharmacist which medicines have a liquid or dispersible form that fits Level 3, ask the GP whether any can stop, and only then consider covert administration for what is left, with a written plan and a review date. The medicines record should then show, for each medicine, the form, the route, whether it is given covertly and the review date. The Mental Capacity Act and DoLS checklist walks through the assessment and best interests steps.

Eating and drinking with acknowledged risks

Eating and drinking with acknowledged risks, sometimes shortened to EDAR, is a formal plan agreed when someone chooses, or it is decided in their best interests, to eat and drink in a way that carries a known risk of aspiration, because the alternative would cost them more than it gains.

It is not a shortcut around a SALT recommendation and it is never a decision one service makes alone. A proper plan involves the SALT, the GP, the person, the family and often a dietitian, records what was discussed, what alternatives were considered, what the person wants and what the service will do to reduce risk while respecting the choice. It has a review date and is revisited after any incident. Where it comes up most is at the end of life, and with people whose distress at modified textures has made meals traumatic. What is not legitimate is a service quietly deciding a level is too much trouble and writing the phrase in the notes afterwards.

What should you record at every meal?

Record the IDDSI level actually served for food and for drinks, the amount eaten and drunk, who supported, the posture and pacing used, any signs observed, and anything left in the mouth at the end. The key field, and the one most services do not have, is the level actually served as opposed to the level prescribed.

A good mealtime record has these fields, and they should be taps rather than typing so they get done:

• Date, time and setting, including whether the meal was at home, at the day service or out
• Prescribed IDDSI level for food and drinks, shown automatically from the profile rather than typed
• Level actually served, selected from the IDDSI list
• What the meal was, in words
• Portion eaten, as a fraction or a simple scale
• Fluid volume in millilitres, and the thickener product and scoops used
• Posture, and whether the person stayed upright afterwards and for how long
• Supervision given: full, line of sight, or none
• Signs observed: coughing, wet voice, throat clearing, pocketing, colour change, distress, none
• Mouth check done and result, where the plan requires one
• Name of the worker, captured automatically

Those fields turn a pile of daily notes into evidence, and they let a manager run a report, which is the difference between recording and knowing. Good care planning systems hold the prescribed level on the profile and pull it into every mealtime entry, so a mismatch between prescribed and served is visible immediately rather than at the next audit.

What should you record after a choking incident?

Record what the person was eating, the prescribed IDDSI level and the level actually served, who prepared the food, who was supervising, what the person was doing, what intervention was used, how long it lasted, the observations afterwards, and who was informed. The texture comparison is the field that makes the record useful.

Every choking event, including near misses the person cleared themselves, should generate an incident record on the day. Back blows or abdominal thrusts mean an ambulance or a same-day medical review, because abdominal thrusts can cause internal injury and because material can remain in the airway. Any event involving loss of consciousness, blueness or a hospital stay is likely to be a CQC notification and a safeguarding referral, and a death is a coroner's matter and a LeDeR notification. Then comes the part services skip: a separate, dated review by the manager asking what texture was served and why, whether the plan was current, whether the worker had been trained, and what has changed as a result. Our guide to incident reporting in care homes covers the review step and the notification thresholds.

Weight, nutrition and hydration alongside dysphagia

Modified texture diets are strongly associated with poor nutrition and dehydration, because pureed food is often served in smaller portions, is less appetising, and loses calories in the liquid added to get the texture right. Thickened drinks are widely under-consumed because people do not like them.

So weigh people monthly, more often if there is concern, and run a MUST score, remembering its limitations where height is hard to measure. Watch fluid totals against a target set with the dietitian, record intake in millilitres rather than cups, and ask about fortification, because a fortified Level 4 meal can carry twice the calories of an unfortified one at the same texture. Our guides to nutrition, hydration and fluid charts and calculating a MUST score cover the mechanics. The link is simple: the safer the texture, the harder it is to eat enough, so the two have to be managed together.

Training: who needs it and how often

Everyone who prepares food, serves food or supports someone to eat needs dysphagia awareness and IDDSI training, including bank and agency staff and kitchen staff. Annual refresher is the usual standard, plus person-specific competency for anyone supporting a person with a known swallowing risk.

The training that changes behaviour involves doing the tests: making a Level 2 drink and measuring it, making a Level 4 puree and doing the spoon tilt, cutting food to 15mm and checking it against the fork. Record the competency, not just the attendance, because a record that the manager watched someone prepare Jordan's lunch correctly on a named date is far stronger evidence than a signed attendance sheet.

Marcus: day services, family meals and the places the plan does not travel

Marcus is 34, has a severe learning disability and autism, and eats Level 6 Soft and Bite-Sized with Level 0 Thin drinks. He lives in supported living, attends a day service three days a week, and goes to his parents' house every other Sunday.

His service got his own house right and lost him everywhere else. The day service had a two-year-old copy of his plan. His mother had never been given anything in writing, only a conversation, and was cutting his food to what she judged was a sensible size. On a trip to a garden centre a support worker bought him a scone.

What fixed it was treating the plan as something that travels: a one-page, dated summary with his photograph, the two levels, the foods to avoid and what to do if he coughs, issued to the day service and to his parents, reissued every time the plan changes with a record of who received it, plus a line in the outing checklist asking whether what he will eat fits Level 6. None of that is clinical, and all of it is the part that actually goes wrong. Services running supported living software should be able to print that summary from the record rather than maintaining it separately, because a separately maintained summary is one that will go stale.

How do you audit dysphagia across a service?

Audit dysphagia by comparing what is prescribed with what is recorded as served, person by person, month by month. Everything else follows from that one comparison, because it is the only measure that shows whether the plan is reaching the plate.

A monthly audit should cover: every person with a swallowing risk has a current SALT recommendation whose review date has not passed; the IDDSI levels on the profile match the letter; the kitchen sheet matches the profile; the thickener product and scoop count match the pharmacy supply; mealtime records exist for the expected number of meals; the level served matches the level prescribed in every record; weights and fluid totals are within range; every choking event has an incident record and a completed review; and every worker on the rota has current training. Turn each into a number and track it across months. A service that can show the mismatch rate falling from four a month to zero is showing something far stronger than a folder of policies.

Common mistakes

• Writing soft diet or thickened fluids instead of the IDDSI number and name.
• Keeping the level in a SALT letter in a folder rather than on the person's profile.
• A handwritten kitchen card that is not regenerated when the level changes.
• Changing thickener brand and keeping the old scoop count.
• Making thickened drinks in advance and serving them after they have stood.
• Recording meals only when something went wrong, so no pattern is visible.
• No field for what texture was actually served, so texture errors stay invisible.
• Crushing tablets at the trolley instead of a pharmacist review of every medicine.
• Letting the level drift upward because the person dislikes the texture, without a SALT review.
• Forgetting the day service, the respite unit and the family.

What good looks like on inspection day

An inspector asks about Jordan. The manager opens his record and the first thing on the screen is Level 5 Minced and Moist, Level 2 Mildly Thick, two levelled scoops of Nutilis Clear per 200ml, full supervision, no bread, no rice, no mixed consistencies, reviewed 6 November, next review in May. The SALT letter is attached and dated, and the kitchen sheet is a printout from that same record, dated this week.

The inspector asks a support worker what they would do if Jordan coughed. They describe stopping the meal, sitting him upright, listening to his voice, recording it and telling the senior, and they say that two coughing episodes in a week would mean a SALT referral. The inspector asks the chef to show the IDDSI test, and the chef takes a syringe and a fork out of the kitchen drawer and does it. The manager then shows the mealtime records for the last month with the level served on every one, the audit showing zero mismatches, and the two near misses in the last year with completed reviews. That is the evidence inspectors look for, and none of it was prepared for the visit.

Final conclusion

Dysphagia in learning disability services is a daily safety task, not an occasional clinical event. Get a real SALT recommendation with the IDDSI number and name for food and for drinks, the thickener product and the exact scoops, the posture, the pacing and the supervision. Put the level where it cannot be missed: the kitchen, the hospital passport, the day service and the family. Teach every worker the syringe test and the fork test. Record the level actually served at every meal, and audit the gap between prescribed and served until it is zero.

Frequently asked

What are the IDDSI levels 0 to 7?

IDDSI runs from Level 0 Thin to Level 7 Regular. Levels 0 to 4 describe drinks: Thin, Slightly Thick, Mildly Thick, Moderately Thick and Extremely Thick. Levels 3 to 7 describe foods: Liquidised, Pureed, Minced and Moist, Soft and Bite-Sized, and Regular or Easy to Chew. Levels 3 and 4 appear in both because the textures are physically the same.

Who decides a person's IDDSI level?

A speech and language therapist assesses the swallow and sets the IDDSI level for food and for drinks. Nobody in the service, including the registered manager, can change it. If the level looks wrong, the service reports the observation, asks for a review and records that it did so.

How do you test whether a drink is the right IDDSI level?

Use the IDDSI flow test. Block a 10ml slip tip syringe, fill to 10ml, release for exactly ten seconds and read what is left. Less than 1ml is Level 0, 1ml to 4ml is Level 1, 4ml to 8ml is Level 2 and 8ml to 10ml is Level 3. For Level 4 use the spoon tilt test instead.

Why is dysphagia more common in people with a learning disability?

Around a third of adults with a learning disability have dysphagia, and more than eight in ten of those with profound and multiple learning disabilities. Cerebral palsy, reflux, poor dentition, posture problems, eating too fast and the sedating effects of antipsychotics and anticonvulsants all contribute, and they usually occur together.

Can a care service change someone's texture because they dislike it?

No. Only a speech and language therapist can change an IDDSI level. If someone is distressed by their texture, refuses meals or is losing weight, that is a reason for an urgent SALT review, not a reason to serve a different level. Record the observation, make the referral and record the date you made it.

What should you record after a choking incident?

Record what the person was eating, the prescribed IDDSI level, the level actually served, who prepared the food, who was supervising, the intervention used, how long it lasted, observations afterwards and who was informed. Anyone who received back blows or abdominal thrusts needs medical review the same day.

Do thickener brands make a difference?

Yes. Scoop sizes and thickening power differ between products, so the same scoop count in a different brand gives a different level. Starch-based thickeners also keep thickening as they stand, while gum-based products are more stable. Record the exact product and scoops per 200ml, and treat a brand change as a change needing SALT and pharmacist confirmation.

What is eating and drinking with acknowledged risks?

It is a formal, multidisciplinary plan agreed when a person chooses, or it is decided in their best interests, to eat and drink in a way that carries a known aspiration risk because the alternative costs them more than it gains. It involves the SALT, the GP, the person and the family, and it is written down with a review date.

Sources

• International Dysphagia Diet Standardisation Initiative (IDDSI) framework and testing methods
• Royal College of Speech and Language Therapists: dysphagia guidance and eating and drinking with acknowledged risks
• NHS England: LeDeR annual reports on deaths of people with a learning disability
• Care Quality Commission: guidance on safe care and treatment and statutory notifications
• British Dietetic Association: guidance on texture modified diets and nutritional adequacy
• Specialist Pharmacy Service: administering medicines to people with swallowing difficulties
• Resuscitation Council UK: choking in adults
• National Institute for Health and Care Excellence: guidance on the care and support of people with learning disabilities