Barthel Index (BI)

Purpose of the measure

The Barthel Index (BI) measures the extent to which somebody can function independently and has mobility in their activities of daily living (ADL) i.e. feeding, bathing, grooming, dressing, bowel control, bladder control, toileting, chair transfer, ambulation and stair climbing. The index also indicates the need for assistance in care.

The BI is a widely used measure of functional disability. The index was developed for use in rehabilitation patients with stroke and other neuromuscular or musculoskeletal disorders, but may also be used for oncology patients.

Available versions
The BI was first developed by Mahoney and Barthel in 1965 and later modified by Collin, Wade, Davies, and Horne in 1988.

  • Original 10-item version (Mahoney & Barthel, 1965). Refers to the following 10 categories: feeding, bathing, grooming, dressing, bowel control, bladder control, toileting, chair transfer, ambulation and stair climbing. Items are weighted according to the level of nursing care required and are rated in terms of whether individuals can perform activities independently, with some assistance, or are dependent (scored as 10, 5 or 0).
Features of the measure

Items of the measure:
The original 10-item form of the BI consists of 10 common ADL activities including: feeding, bathing, grooming, dressing, bowel control, bladder control, toileting, chair transfer, ambulation and stair climbing. Items are rated in terms of whether individuals can perform activities independently, with some assistance, or are dependent (scored as 10, 5 or 0). Items are weighted according to the level of nursing care required.

 Scoring:
The score of the BI is a summed aggregate and there is preferential weighting on mobility and continence. The scores are allotted in the following way: 0 or 5 points per item for bathing and grooming; 0, 5, or 10 points per item for feeding, dressing, bowel control, bladder control, toilet use, and stairs; 0, 5, 10, or 15 points per item for transfers and mobility. The Index yields a total score out of 100 – the higher the score, the greater the degree of functional independence (McDowell & Newell, 1996). This score is calculated by simply totaling the individual item scores, which requires simple arithmetic computation
by hand.

A modified scoring system has been suggested by Shah, Vanclay, & Cooper (1989) using a 5-level ordinal scale for each item to improve sensitivity to detecting change (1=unable to perform task, 2=attempts task but unsafe, 3=moderate help required, 4=minimal help required, 5=fully independent). Shah and coll. (1989) note that a score of 0-20 suggests total dependence, 21-60 severe dependence, 61-90 moderate dependence and 91-99 slight dependence.

Subscales:
None typically reported.

 Equipment:
To administer the BI, one only needs a pencil and the test items.

 Training:
Administration of the BI does not require training and has been shown to be equally reliable when administered by skilled and unskilled individuals (Collin & Wade, 1988). The BI can also be self-administered (McGinnis, Seward, DeJong, & Osberg, 1986). However, for patients older than 75 years of age, it is not recommended that the BI be administered as a self-report measure (Sinoff & Ore, 1997). One study suggests that the scale can be administered reliably over the telephone (Korner-Bitensky & Wood-Dauphinee, 1995).

 Time:
The BI can take as little as 2-5 minutes to complete by self-report and up to 20 minutes to complete by direct observation (Finch, Brooks, Stratford, & Mayo, 2002).

Alternative forms of the BI
    • Modified 10-item version (MBI)(Collin et coll., 1988).Functional categories may be scored from 0 to 1, 0 to 2, or 0 to 3, depending on the item. Total scores range from 0 to 20.
    • 5-item short form(Hobart & Thompson, 2001).The 5-item version refers to the following 5 categories: transfers, bathing, toilet use, stairs, and mobility. Each item is scored 0 to 1, 0 to 2, or 0 to 3, depending on the function. Total scores range from 0 to 20. Hobart & Thompson (2001) found that the 5-item BI is psychometrically equivalent to the 10-item BI (correlation with original version was r = 0.90).
    • The expanded 15-item version(Granger et coll., 1979; Fortinsky & Granger, 1981). Added a 4-point scale of intact/limited/helper required/null. Scores range from 0 to 100. In the 15-item version, a score of 60 is commonly considered to be the threshold score for marked dependence (Granger, Sherwood, & Greer, 1977). High correlations of the expanded 15-item BI and other measures of function have been demonstrated (e.g., with Katz Indice of Activities of Daily Living, r = 0.78; with PULSES profile (medical status, upper and lower limb function, sensory and excretory function, mental and emotional status), r = -0.74 to -0.90 (Shinar, Gross, Bronstein, Licara-Gehr, Eden, Cabrera, et coll., 1987; Granger, 1985; Rockwood, Stolee & Fox, 1993). Scores were also predictive of return to independent living after 6 months (Granger, Hamilton, Gresham, & Kramer, 1989).
    • The extended BI (EBI)(Prosiegel, Bottger, & Schenk, 1996). The EBI consists of 16 items, 15 of which are identical to the Functional Independence Measure. Very little literature exists on the EBI, however Jansa, Pogacnik, and Gompertz (2004) found it to be a reliable and valid measure of disability/activity levels in 33 patients with newly diagnosed acute ischemic stroke.
    • The 3-item BI(Ellul, Watkins, & Barer, 1988).Based on 3 items (bed-chair transfers, mobility, and bladder incontinence), it is a useful alternative to the full BI for assessing function at hospital discharge. To date, this version has only been validated in patients with stroke.
    • Self-rating BI (SB). The SB has good concurrent validity and is well related with the original BI and the Functional Independence Measure. The indexes test-retest reliability is sufficiently high for practical use (Hachisuka, Ogata, Ohkuma, Tanaka, & Dozono, 1997; Hachisuka, Okazaki, & Ogata, 1997; McGinnis et coll., 1986).
    • Early Rehabilitation Barthel Indice (ERI). An extension of the BI, it was developed to assess functioning of individuals with severe brain damage, who often cannot be differentiated appropriately due to floor effects that occur with increasing severity of neurological impairment. The ERI looks at the following aspects: state requiring temporary intensive medical monitoring, tracheostoma requiring special treatment (suctioning), intermittent artificial respiration, confusional state requiring special care, behavioural disturbances requiring special care, swallowing disorders requiring special care, and severe communication deficits. Schonle (1995) found that the ERI is quick, economical, and reliable when administered to 210 early rehabilitation patients and 312 patients with severe brain damage.

    There is little consensus over which should be considered the definitive version of the BI (McDowell & Newell, 1996), but the original and the 10-item and 15-item modifications are the most commonly used.

Client suitability

Can be used with:

Patients with stroke.

The BI is a frequently used stroke outcome measure. It has been repeatedly shown to be a reliable and valid measure of basic Activities of Daily Living (Mahoney & Barthel, 1965; Loewen & Anderson, 1990; Gresham, Phillips & Labi, 1980; Collin et coll., 1988; Roy, Tongeri, Hay, & Pentland, 1988; Wade & Hewer, 1987; Leung et coll., 2007). In patients with stroke, the BI determines the extent of post-stroke disability, self-care activities and ability to live independently. The total score of the BI has also been found to predict length of stay in hospital (Granger, Albrecht, & Hamilton, 1979).

There are no prerequisites for completing the BI. For patients who are unable to respond to the BI independently, the BI can be completed by proxy (eg. Duncan, Lai, Tyler, Perera, Reker, & Studenski, 2002; Wyller, Sveen, & Bautz-Holter 1995). Further, the BI can be reliably administered over the telephone to either the patient or their proxy (Korner-Bitensky & Wood-Dauphinee, 1995).

 Should not be used in:

  • To capture significant losses in higher levels of physical function or activities that are necessary for independence in the home and community. This means that patients can still score a maximum score of 100 and experience significant impairments
    (Kelly-Hayes et al., 1998).
  • It should be used with caution in patients with mild stroke. It is responsive to change but has definite ceiling effects in persons with mild stroke (Wade & Hewer, 1987; Skilbeck, Wade, Hewer, & Wood, 1983).
In what languages is the measure available?

 The BI has been translated and validated in:

  • Dutch (Post, van Asbeck, van Dijk, & Schrijvers, 1995)
  • German (Heuschmann et al., 2005; Valach, Signer, Hartmeier, Hofer, & Steck, 2003)
  • Turkish (Kucukdeveci, Yavuzer, Tennant, Suldur, Sonel, & Arasil, 2000)
  • Persian (Oveisgharan, 2006)
  • French (Condouret et al., 1988; Wirotius & Foucher-Berres, 1991)
  • Chinese (Leung, Cha, & Shah, 2007) (modified Barthel Index)