Average length of stay (ALOS): This measure refers to the average number of days that a patient stays in a hospital. It is calculated using the following formula:

ALOS= (Inpatient days)/Admissions (1)

Bed occupancy rate (BOR): The occupancy rate is a measure of utilization of the available bed capacity. It indicates the percentage of beds occupied by patients in a defined period of time, usually a year. It is computed using the following formula:

BOR= (Inpatient days)/(Bed days) ×100 (2)

Where,

Inpatient days = admissions × ALS, and bed days in the year = number of beds × 365

Bed turnover ratio (BTR): The turnover ratio is a measure of productivity of hospital beds and represents the number of patients treated per bed in a defined period of time (usually a year). It is computed as follows:

BTR= (Total patient admissions)/(Number of beds) (3)

The three performance indicator ratios commonly used in assessing the relative performance of hospitals including:

- Average length of stay (ALOS): This measure refers to the average number of days that a patient stays in a hospital. It is calculated using the following formula:

ALOS = Inpatient days / Admissions                                                            (1)

- Bed occupancy rate (BOR): The occupancy rate is a measure of utilization of the available bed capacity. It indicates the percentage of beds occupied by patients in a defined period of time, usually a year. It is computed using the following formula:

BOR= Inpatient days / Bed days ×100                                                        (2)

Where,

Inpatient days = admissions × ALOS, and bed days in the year = number of beds × 365

-Bed turnover ratio (BTR): The turnover ratio is a measure of productivity of hospital beds and represents the number of patients treated per bed in a defined period of time (usually a year). It is computed as follows:

BTR = Total patient admissions / Number of beds                                       (3)

Dear Younsi Moheddine, thank you for your time helping with my question.

I think I shall apologize for the poor phrasing of my question, that may have led you to mistook my question above.

I am familiar with the formulas you have kindly listed above, however I am not very sure how these indicators should be reported in term of a country or a state (not an individial hospital).

Shall I be taking the whole country as one unit (or taking a state as a unit) when reporting BOR/ALOS/TOI for a country ? With that I mean, should I be (take BOR for example) adding all patient days for all hospitals within a country and divides it by the total number of bed days for all hospitals for the respective country ?

Or shall I be taking the average BOR of all hospitals within the country, and thus obtaining the "Average BOR" for the country(or state)?

Reading OECD Health Statistics, I am not sure which way are they reporting, thus when calculating and comparing my country statistics, I face this problem.

Hope you may assist me with your expertise and experience.

Thanks

Thanks James for your kind input.

We have this very similar situation : small hospitals with extreme statistics skewed the overall statistics. We were not sure how other countries report their statistics. We worry that our statistics may not be comparable if calculated differently from others, thus thought of getting opinion from researchers with similar experience.

We have finally decided to report these statistics by taking a state (or a country) as a unit instead of averaging the hospitals' statistics.

I have tried searching for literature as guide, have yet found any useful ones. Any pointer?

Hi

I would be  concerned  regarding reporting the  "Average LOS" only   for  a hospital and  or  a state .

However, Caution might need to be practiced when reporting the "mean"  LOS only , since  many time-bound  process,  human performance measures  related to time  are always and -not necessarily always -  skewed. For Length of  Stay measures based on monthly, quarterly however they'd come  skewed  very often. Reporting the median, IQR along with the mean can be  helpful. The reason behind this is that  the IQR and median are quit immune against outliers or extremes  compared to the mean as a statistic . Also, I would  gather   monthly LOS  measures  across time -or simply  through historic pool ( Say 30 and more months) and  do  Modeling for the  LOS  before  deciding  on a benchmark, especially for  hospital systems which has just started  reporting  LOS as a quality Indicator. Worthwhile  also, to identify any seasonal  variations,  possible   special causes for either drop or an increase on LOS  in historical  data. This, however, can  help clarify   many   shifts in  individually computed  LOS ( say Monthly ). An example   to make this  in simpler construct , in Saudi  Arabia  and particularly where I work, there  is   trend in   numbers of admissions per month  that  keep cycling  nearly every year  ( higher admissions  that stay longer  during  wind and sand storm seasons: Asthmatic , COPD,  the Immuno-compromised patients can suffer from more  sequel  due to cold and  sand storms ). Although this can be accounted  for statistically using adjustment , but requires  someone to  be statistically vigilant and maybe requires the hospital to  incorporate  risk  scoring systems into their  units,  which might take  a lots of , time ,work and effort.

So, In simple : I would  keep an eye at  mean, median and IQR, Minumum, maxixmum  LOS Not only the mean ( Average) LOS.

I hope this  resource would be of use:

http://www.phc4.org/reports/hpr/98/techguid/length_of_stay_analysis.htm 

 best regards,

 Mohammad

Several useful papers to help are available at: www.hcaf.biz in the Hospital Beds web page

The metrics mentioned above are derived metrics and should not be summed up. Instead, the individual components need to be summed up to calculate the metric for the combined entity. E.g. To calculate average ALOS across a group of hospitals, add up the total inpatient days for all hospitals and divide by the total admissions for all hospitals. Similarly, for occupancy, add up the inpatient days for all hospitals and divide by the total bed days for all hospitals. 

how can we calculate number of bed days in a hospital?

Dear Sajid

if you send me an email contact I will send you documents with the formulas embedded as we used them in the Hospital in which I was a nurse manager.

I need to hunt them out of my archives.

Jane

[email protected]

 Dear Jane Kerr

I hv sent email contact but u still didnt send those formulas. kindly send me as soon as possible

I am trying to rank hospitals by ALOS, between 1.9 and 37, which one is the most efficient, is it 1.9 or 37. and going by that should I rank them in ascending or descending order

A lot depends on how many same day stay admissions are included in the average and if LOS for each patient is calculated as a whole number or in days hours and minutes. Whole hospital LOS needs to be adjusted for differences in case mix, i.e. a maternity hospital will have a lower LOS than a specialist cancer hospital, etc. See http://www.hcaf.biz/2010/Publications_Full.pdf for useful papers regarding LOS comparison.

This paper illustrates many of the issues.

http://www.hcaf.biz/2018/Maternity_LOS.pdf

Comparing the BOR between hospitals and as country averages is fraught with danger. The BOR depends on hospital size. Queuing theory sheds light on this relationship. The best way to compare BOR is to convert it into turn-away, i.e. the percentage of time that a bed is not available for the next arriving patient. See discussion documents on the right hand side at http://www.hcaf.biz/hospitalbeds.html

Do not get fooled into thinking that BOR is a measure of efficiency, the exact opposite is the case.

I have question from experts here, that I guessed more or less related to our discussion. I'm quite confused with the terms should be used for them..e.g. "patient discharge"/ patients admission/readmission.

What is kind of category they are belong to: - Healthcare/Hospital delivery process, -healthcare demand , -healthcare service utilization, healthcare capacity/resources utilization or etc...

If possible, may you share here a great sources that explain all these terms..

Thank you =)

Hi Fatin,

More traps for the unwary. So-called admissions are usually counted at the point of discharge! Hence if you want a count of admissions per month you need to ask for exactly that or else it will be assumed that you want a count of discharges.

Re-admissions are when the same person get re-admitted usually within 30 days of the discharge. However, be aware that 50% of a persons lifetime use of acute inpatient care occurs in the last 6 month of life, hence, the re-admission rate is most probably a measure of the number of persons in the last 6-months of life rather than of hospital care. See references at http://www.hcaf.biz/2010/Publications_Full.pdf

Happy learning. Cheers Rod

While slightly diverging from the original question, this short paper may challenge your perceptions regarding the 'robustness' of hospital bed planning and forecasting. Happy cogitating. http://www.hcaf.biz/2018/Beds_Disease_X.pdf

The gold standard for calculating LOS is to use days, hours and minutes. My own studies show that the real average LOS can be underestimated by something in excess of 4.5% by the use of integer values (usually a count of midnight stays). A useful reference is Tierney L, Conroy K. Optimal occupancy in the ICU: A literature review. Australian Critical Care 2014; 27: 77-84. Available via Research Gate Article Optimal occupancy in the ICU: A literature review

As you will all know the bed occupancy rate varies over time. I have proposed that this is largely due to changes in the number of deaths, since the last year of life (especially the last 6 months) is the most hospital intensive. To illustrate this concept I have demonstrated large changes in the ratio of healthcare staff per death in England, see http://www.hcaf.biz/2018/Staff_Death.pdf Hopefully this will add to the discussion around factors regulating bed demand and occupancy.

The ratio(s) of available and occupied (midnight) beds per death is a useful way to compare different healthcare systems, see http://www.hcaf.biz/2018/Beds_per_death.pdf This short paper looks at the trends in beds per death in England since 1987. Hope this is useful.

For those interested in my full series of articles on hospital bed numbers and occupancy please refer to http://www.hcaf.biz/2010/Publications_Full.pdf where a draft version of many of the papers can be downloaded. Please publish your own studies in this much neglected area of healthcare planning. Always happy to answer questions at [email protected]

If you wish to compare available beds between countries see the supplementary material in http://www.hcaf.biz/2018/International_Beds.pdf

Recent research has demonstrated that the nearness to death effect has profound implications to acute inpatient demand. The central issue is that 55% of lifetime acute bed occupancy occurs in the last year of life. Hence the absolute number of deaths drives the marginal changes in bed occupancy. Deaths are very volatile from one year to the next and this then drives the volatility in bed demand and hence the need for a generous occupancy margin. Please investigate these issues and publish your findings. See http://www.hcaf.biz/2010/Publications_Full.pdf for associated studies relevant to on/off switching in deaths, etc.

Clearly, it does not seem a wise use of acute resources for 55% of a persons lifetime acute care in the last year when death is the only eventual outcome. Models are needed to detect which persons are in the last year of life to deliver care outside of an acute setting.