Title: Economic Costs of Dysphagia Among Hospitalized Patients
Authors: Allen, Greene, Sabido, Stretton, & Miles
Journal: The Laryngoscope
Year of Publication: 2019
Design Type: Retrospective observational cohort study
Purpose: “This study explored the additional burden of a swallowing problem on patients admitted with a medical event, not typically associated with new-onset of swallowing difficulties—hip/femur fracture.”
Population: adult patients admitted to a hospital for hip/femur fractures with and without identified dysphagia
Inclusion criteria: hip/femur fracture diagnosis; pre-existing conditions; “additional stays in rehabilitation wards were part of this analysis”
Exclusion criteria: non-hip/femur fracture diagnoses
Who here likes to save money? I know, duh, right.🙄
Answering that one question is pretty quick and easy for most people, but the authors did their duties to really analyze and break down just what we are doing and how much it costs🤑:
“In the community, these individuals may make adjustments to their diet, use eating strategies to manage food intake, or be able to deal with small amounts of misdirected food and fluid without succumbing to illness. However, in the setting of an additional impairment, particularly an acute change or one requiring surgery, swallow problems may be destabilized and have a negative impact on a patient’s outcome, such as lengthening hospitalization days, increasing hospital readmissions, and increasing need for long-term rehabilitation care.” p.1
There’s gonna be a lot to unpack here, so let’s start with a few basics so we know just what this study did:
“Retrospective” = looks backward in time and examines factors that could lead to an already established outcome before the study
“Observational” = no changes outside of the daily “normal” way of doing things (remember, this simply describes things and canNOT say anything about cause/effect relationships)
“Cohort” = a group of people that usually have the same ‘trait’ or characteristic
Data from 2,453 patients at a single hospital were analyzed between 2015-2018. But they didn’t just look at anyone, the authors specifically picked out those who were admitted year after year with a primary diagnosis of acute hip/femur fracture.
Wait, why are we talking about hip/femur fractures and swallowing??????
Well, my dear friends, many of us SLPs understand that if someone is admitted for something that is directly related to dysphagia, like stroke, for example, we are absolutely going to need to see them for services–be it a screen, clinical bedside evaluation or instrumental, a follow-up, or intensive treatment. So, we need to test the same hypothesis, “do patients with dysphagia have an extended length of stay (LOS) and increase financial burden?” to a different group that is admitted for other causes. That way, we can compare something we absolutely know to be true to be able to get some answers about something we don’t know much about.
So to get more answers, the authors divided the targeted population into 2 groups to compare:
- hip/femur fracture with dysphagia (HF+D) [n=165]
- hip/femur fracture without dysphagia (HF-D) [n=2,288]
A quick side point here:
The authors (and really, hospital analysts🤓) used the coded data from the hospital’s database system and specifically were looking for certain codes based on the ICD-10 (*Australian Modification)..unfortunately, no specifics for which codes were provided.😉
🙌THIS🙌 guys is why our documentation can really matter! Believe me, it’s still a hard mountain that I’m still climbing as well to get that “perfect note”😓. But one way or another, what we say in our documentation, when and how we say it, and why we are clicking on that dysphagia diagnosis box to attach to a patient’s permanent chart will matter in the end. Whether it be for later review on a patient’s lifelong medical record, or some research study that will use that data at some point, it matters.😉
The authors and “experienced otolaryngologist fellows familiar reading clinical notes” collected additional clinical and demographic data to tease out later as well so we can know if any of this made a difference:
- postoperative factors such as Length of Stay (LOS)
- referral to SLP for a swallowing assessment
- type of SLP intervention
When it related to cost, with the help of trained hospital analysts, they were able to further look into:
- the complete episode of care (standard charges for board&care, surgical intervention, and staffing for initial acute admission)
- (the charges are initially in New Zealand dollar+taxes, sorry guys feel free to do the conversions💸💶)
- readmissions within 28 days to the same district health board
- mortality within 30 days following the index admission (taken from national patient ID database)
- a separate analysis computer-randomized for 100 patients from each group was also completed
Now, this is where what some people might call ‘fun,’ begins🤪. No, it’s not exactly my ideal fun-filled day of pizza, puppies, and the beach. But it is sometimes helpful for ME to remember this stuff, and if not fun for you (which I 110% understand😂), the best I can hope for is maybe, informational?🤷♀️
Let’s talk statistics.
Wait!! Come back! I know, these things do 🙅♀️NOT🙅♀️ get me jumping off the walls either (I’m just not a numbers girl unless it’s related to my pizza order🍕). The authors directly describe the type of statistics they used for their data analysis, and if you’re anything like me, you skim over these about as fast as I try to forget the number of cups of coffee or wine I’ve already had🙃… But this time, I really found myself wanting to know the work behind what we’re reading and try to learn what exactly those things actually mean, and more importantly, if and why they matter? (I also will not take it personally if you wanna skip this part😅, maybe just plan on coming back to it later🤞).
To ease into it:
Descriptive statistics= summarizes and describes what the numbers show without making any comparisons
- ex: “the total number of apples was 29” “the mean grade was a B+”
- ‘average age of HF+D group was 85 years/ average age of HF-D group was 78 years‘
categorical variables= words that represent categories
- ex: disease types, gender, pain rating scale
- HF+D, HF-D
continuous variables= can be infinitely divided further into increments
- ex: height, time, distance
- Age, LOS
linear regression= used when we want to predict a relationship between 2 or more groups in a linear way
- ex: if 1 goes ⬆️ the other goes ⬆️ ; if 1 goes ⬆️the other goes ⬇️ etc.
- “Because age differed significantly between the two groups, further linear regression analysis was conducted to control for age, comorbidities, and aspiration.” (p.2)
independent t-tests= used to compare 2 groups to see whether the groups are actually statistically different or not (ex: HF+D compared to HF-D)
Ok, I think that’s enough for today..
Yes, that was a deep brain dive into some nerdy stuff🥴. But, the silver lining lesson I’ve been starting to learn and why it can matter to us clinicians who are arm-deep in barium or thickener, is that sometimes researchers and research studies do NOT use the correct data and statistical tests based on what they are looking at and trying to find.🤯
Why can that matter to us clinicians?🤔
Because yes, we don’t have the time to check all the statistics in our brains (nor would we want to?). And yes, we clinicians are spending almost every minute of our workday (and sometimes before/after that!) focusing exclusively on our patients, how they’re doing, what we can do to help them progress, how we can make an impact in some way for them and their families with very little time to even find a quiet moment in a bathroom (hopefully not to just cry), let alone check to make sure research is using the correct statistics!!
But, unfortunately, this happens and it can have an impact on our clinical practice, which then impacts the progress our patients make. So while yes, we have a LOT on our plates as clinicians, and while memorizing all the statistical tests that go with what data would be the most unrelatable and absurd solution, what we can try to do is make little improvements along the way and remember that while clutching onto the evidence we hold so dear, we also need to keep a critical eye🧐to avoid praising a study that may claim to give us the ‘answer to our problems,’ and forgetting about one that while not 100% significant, at least has a true finding we can rely on.
Onto what they actually found. . .
But first, what do YOU think the results would be for all the factors listed above???🤔
Do you think a dysphagia diagnosis won’t matter as much for those admitted with hip fractures?
Do you hypothesize a difference in who pays more??
Overall, out of 2,453 patients that were analyzed, the group that was identified having dysphagia (HF+D) resulted in:
- higher LOS
- higher costs$$$
- higher mortality
- higher comorbidities
- **regardless of age
Suffice it to say, this group with dysphagia present stayed longer, were worse off medically, and paid more.
“Length of hospital stay was significantly longer in those patients reporting dysphagia compared with those without eating issues (32.3 days +dysphagia vs 14.1 days –dysphagia), and as a result, cost of hospital stay was also significantly greater in the HF+D than in the HF-D group”
“The inpatient LOS more than doubled from 14 to 32 days, reflected in a NZD$14,691 (USD$9,975) increase in cost of admission”
“The mortality rate (within 30 days of discharge) was also substantially greater at more than four times the rate of those without swallow issues…4% in the HF-D group versus 18% in HF +D…and rate of aspiration pneumonia was 14-fold greater in swallow impaired patients compared to normal swallowers (9.7% vs. 0.7%, respectively).” p.2
Here’s how the other factors broke down:
LENGTH OF STAY:
- dysphagia diagnosis = +2.95 days
- > 80 yrs = +5.95 days
- comorbidity = +2.7 days (per each one)
“Adjusting for age, comorbidity and medication still demonstrated dysphagia to be associated with increased cost, mortality, and increased rate of aspiration pneumonia.” p.3
- Dysphagia diagnosis alone = added $3,374NZD (New Zealand dollars)
- age >80 yrs. = added $3,063 NZD
- per comorbidity = added $2,091 NZD each
- presence of dysphagia = increased risk of aspiration pneumonia (despite age/comorbidities)
- dysphagia + age >80 yrs + 5+comorbidities (excluding dysphagia) + aspiration pneumonia + mortality in the hospital→ suggested a very strong predictor of cost burden (also strongly suggesting these determine LOS and driving costs upwards↑)
I think here is where we can ask ourselves, which of these results are surprising?🤨
Which of these could we have intuitively assumed??🧐
More importantly, which of these, might we be overlooking???🤔🤔
To try to sum this all up:
If a patient is 80+ yrs old and admitted for an acute hip fracture:
They’re more like to have dysphagia, which puts them more at risk for aspiration pneumonia and mortality in general, and they’re also more likely to have more comorbidities and medications, stay significantly longer in the hospital, and pay more for it all.😬😳😓🤯
Still have some unanswered questions???
Me too!! Like,
Does this mean we have to assess every single patient post-hip fracture?
Well, no. And we also can’t necessarily generalize this to all populations. The authors even acknowledge the fact that even though the sample size is large (yay!), miscoding errors obviously could occur; but apparently, more than anything this means they would likely be underreported versus overreported! While I may be the only one who’s super interested in the long-term impact of the study’s findings, we also can’t generalize this to the public hospital setting for long-term policy guidance. A final thought that came up for me was how the ‘dysphagia code’ was actually added, clinical or instrumental evaluation??🤷♀️
In the bigger picture, if a patient does have any of the above characteristics, you can certainly bet on a needed consult✅.
Does the hip fracture cause dysphagia?
Well, let’s not jump too fast with that conclusion either..🤨 Especially because this all still needs to be studied further and currently is only qualified for “Level 3b” evidence and further studies definitely still need to investigate more, if there are any takers:😉
“An intervention study is required to evaluate this approach in the public hospital setting and provide long-term policy guidance.” p.6
But given what we know with age, with comorbidities, and especially with muscle atrophy, deconditioning, medication interactions, and sometimes hospital delirium, you can also bet any one of these can send someone’s swallowing ability in a snowball spiral:
“If a major event occurs, such as a traumatic hip fracture, the preexisting swallow impairment may be exacerbated by deconditioning, need to remain nothing per mouth pending surgery, anesthetic instrumentation, difficulties with mealtime positioning, acute changes in cognition, and additional medications given during hospital stays.” p.4
With this specific elderly population we can hypothesize that there may be some underlying dysphagia or presbyphagia concerns, on top of the acute diagnosis like hip fracture:
“Although some individuals may experience a precipitous deterioration in health during admission and thus only begin to experience swallow decline once in the hospital, it is more likely that most of the case cohort (HF+D) presented some swallow deficit at admission.” p.5
“Attention may focus on the acute injury, and caregivers may fail to identify further decline in swallow function while the individual is experiencing pain or distress.”
“Only a small proportion of those with swallow deficits seek and receive specific treatment for their symptoms, which often coexist with other disorders and can be affected by medications, general health, and diet” p.4
Where to go from here?
Just in case my wrapping-it-up skills didn’t do the trick, the authors tie all this up in a nice bow for you and even give some advice for 🚩red flags🚩:
“We believe these data offer guidance as to what risk factors may indicate the need for swallow assessment if recorded at admission. Analyses supports older age (particularly age over 80
years), American Society of Anesthesiologists classification, more than eight pre-existing additional comorbidities, more than five pre-existing regular medications, and known diagnosis associated with dysphagia as red flags.
These patients should be referred for full speech language therapy evaluation and observed closely for signs of pulmonary problems and nutritional deficits.” p.5
and even some resources:
“Simple screening tools are available and have been validated to detect swallow risk, nutritional risk, and frank malnutrition, including the Eating Assessment Tool-10 and the Mini Nutritional Assessment-Short Form. Tools such as these may be employed to identify risk and which patients should be referred for speech therapy review, and enable dietary adjustments to be made to suit the individual.” p.5
(Might as well toss in Yale Swallow Protocol for aspiration just in case😅💁♀️. Also, I believe I may have heard recently that there might be a newer and better tool than the EAT-10, specifically for the HNC population? Anyone feel to correct me on this!👍)
How can this help YOU????
I for one, think this article is anything but simple and not limited to just “hip fractures and dysphagia.” There are a lot more questions to think about that ultimately (I think) relate to discovering our real roles as SLPs, as well as finding the right balance between:
- over-diagnosing vs underdiagnosing dysphagia
- providing skilled vs non-skilled services (and also the needed duration of those services😉)
- doing whatever is necessary to prevent negative outcomes for the patient vs also recognizing the patient’s financial burden
- Do you think this article could be helpful to advocate for patients, caseload, and ST services????
- Do you think this could be something to bring up in physician rounds?
- Should we be sharing that list of hip surgery patients w/PT/OT??
- How closely should we be an automatic referral for this population, instead of just stroke???
- “In a recent systematic review of 34 studies, patients complaining of dysphagia incurred a 40% increase in costs during their hospital stay compared to those without, and in a 1-year follow-up of more than 3,000 stroke patients comparing costs in those with dysphagia complaints versus those without, there was a USD$4,510 increase in expense in those reporting swallow complaints”
- “The effects of a major injury such as hip fracture may compound preexisting swallowing dysfunction or destabilize deglutition enough to overwhelm physiological reserve precipitating aspiration or pneumonia. Significantly higher prevalence of more than five preexisting comorbidities and more than five preexisting regular medications in those with swallowing complaints backs up this theory.”
- “Despite being coded with dysphagia on discharge, and struggling with eating and drinking while on the orthopedic ward, less than half of patients were referred to speech-language pathology, and only a third received a dietitian review. Referrals were often delayed, with most not occurring until week 3 of admission.”
- “Routine screening for swallow or nutrition status is not conducted but should be considered. Instituting screening processes to identify impaired swallowers early in their admission may provide an opportunity to intervene positively in this process.”
- “If swallow impairment is recognized at admission or early in a hospital stay, simple strategies can be employed to decrease the risk of mismanagement of food and fluid. Speech-language pathology intervention, dietary advice, and assistance during mealtimes provide substantial protection from choking, misdirection of the bolus, and pulmonary soiling.”
Allen, J., Greene, M., Sabido, I., Stretton, M., & Miles, A. (2020). Economic costs of dysphagia among hospitalized patients. The Laryngoscope, 130(4), 974-979. doi:10.1002/lary.28194