Circumferential Measurements to Calculate Lower Limb Volume in Persons with Lymphedema: What Segment Length Is to Be Recommended?

Abstract

Background:

Circumferential measurements (CMs) every 4th cm are commonly used to assess lower limb volume (LLV), but fewer measurements would be less time-consuming. The aim of this study was therefore to establish the agreement between LLV measurements derived from CM every 4th cm (V4), 8th cm (V8), and 12th cm (V12), and to evaluate the intrarater test–retest reliability for each of the three measurement methods in persons with lower limb lymphedema (LLL).

Methods and Results:

Forty-two persons with unilateral or bilateral LLL were measured twice, 2 weeks apart. Volume measurements for the V4, V8, and V12 methods were derived using CM. The agreement was evaluated using intraclass correlation coefficient (ICC_3.1) and Bland–Altman graphs including 95% limits of agreement (LOA). The reliability was evaluated using ICC_2.1 and standard error of measurement (SEM%) and smallest real difference (SRD%). The agreement was high for the V4 and V8 methods (ICC 0.999), and for the V4 and V12 methods (ICC 0.998). The graphs revealed slightly higher agreement between the V4 and V8 than between the V4 and V12 methods visualized by the 95% LOA (−117 to 62 and −236 to 132 mL, respectively). For all three measurement methods, the test–retest reliability was high (ICC 0.993–0.995) and the measurement error low (SEM%: 1.2%–1.4% and SRD%: 3.4%–3.8%).

Conclusions:

The higher agreement between the V4 and V8 methods than between V4 and V12, and the high test–retest reliability in LLV measurements support the V8 method to replace the V4 method in persons with LLL.

Background

Lymphedema (LE) is considered a chronic disease characterized by increased volume of the affected limb or limbs.¹ To measure lower limb volume (LLV) and changes over time is therefore essential in persons at risk of or diagnosed with lower limb lymphedema (LLL). Reliable measurements of LLV will help clinicians to diagnose LLL, determine the stage of LE, plan appropriate management, and evaluate effects of treatments.¹

Various measurement methods can be used to assess LLV. The water displacement method (WDM) is standard in upper limb LE,² but is not so common in LLL due to bulky equipment, the large amount of water needed, and the extensive cleaning up efforts after use. The advantage with WDM is that the measurements include the entire limb volume including the foot. A disadvantage, however, is that there is limited knowledge about the reliability of LLV measurements.³

The Perometer, an optoelectronic measurement method for LLV measurements, has a short measurement time^4,5 and a high test–retest reliability in healthy persons.^5,6 The disadvantages are that the Perometer is expensive and mostly used in specialist clinics and there is lack of knowledge about the reliability for LLV measurements in persons with LLL.

The tape measurement method that is standard for volume measurements in LLL^1,2 and circumferential measurements (CMs) every 4th cm along the limb is commonly used.^2,5,7,8 The advantages of this method are the low cost, it is easy to clean, and requires limited space. Disadvantages are that CMs every 4th cm is time-consuming, and that the measurement procedure could be challenging (i.e., positioning of the first measurement point, choosing the right angle for the tape on the limb, and tension on the tape).⁹ However, by using a standardized measurement procedure, reliable CMs with small measurement errors can be obtained.^7,8,10

To use fewer CMs than every 4th cm for LLV measurements would be desirable as this would save time in the clinic. Currently, few studies have evaluated how many CMs are needed for reliable assessments of LLV. Tidhar et al¹¹ evaluated test–retest reliability of LLV measurements derived from CMs at only 8 predefined points in five persons (two healthy and three with LE) and found that standard error of measurement (SEM) was low. Mayrovitz et al¹² investigated therapy-related changes in LLV measurements based on CMs every 4th cm, 8th cm, and 12th cm and found no difference in volume reductions regardless of the method used. However, when investigating if a new measurement method can replace an old one,^13,14 several statistical methods are recommended such as calculating agreement, mean differences, and measurement errors.¹⁴

Therefore, the aim of this study was to: (1) establish the agreement between LLV measurements derived from CMs every 4th cm (V4 method; reference standard) with CMs every 8th cm (V8 method) and every 12th cm (V12 method), and to (2) evaluate the intrarater test–retest reliability for each of the three methods in persons with LLL.

Methods

Research design

A test–retest design was used in the present study.

Participants

Between April 2018 and March 2019, 42 persons with LLL were recruited from the LE unit at Skåne University Hospital. The inclusion criteria were as follows: (1) 18 years or older; (2) a diagnosis of unilateral or bilateral primary or secondary LLL (assessed by lymphoscintigraphy and/or a medical specialist); (3) persistent LE for the last 6 months; (4) a total limb volume variation ≤5% for each limb the last 6 months; (5) treatment with compression stockings daytime or day and night according to usual care; and (6) a compression garment not older than 2 months at the time of inclusion.

The exclusion criteria were as follows: (1) comorbidity such as heart failure, kidney disease, or venous insufficiency that could affect swelling of the LLs; (2) prosthetic knee or hip implants; (4) muscular disorders of the LLs; (5) intake of diuretic medication or any other drug that may interfere with the volume of the LLs; and (6) inability to understand written or oral information in Swedish.

Ethics

Before inclusion, all participants received written and oral information about the study and provided written consent to participate. The study was approved by the Regional Ethics Committee Review Board in Lund, Sweden (Dnr 2016/136).

Measurements

Body mass index was calculated (kg/m²) using the weight measured on a digital scale and the body height reported by each participant.

Experience of heaviness and tightness in the LE limb/limbs over the past week was rated using a 100-mm visual analogue scale¹⁵ ranging from “no discomfort” (0 mm) to “worst imaginable discomfort” (100 mm).^16–18

Thickness of the subcutaneous tissue¹⁹ of the LLs was assessed with the subject in the supine position with bent knees. Palpation of the tissue using the thumb and index finger was performed at the dorsal, lateral, and medial side of the lower part of the limbs and lateral, anterior, and medial side of the upper part of the limbs.^8,10 Increased thickness was noted as yes or no.

Leisure time physical activity during the last 6 months was rated using a six graded classification system,²⁰ ranging from very low to regular/very strenuous activity. This classification system has been validated for a Scandinavian population.²⁰

Lower limb volume was calculated using the tape measurement method and CMs every 4th cm (V4 method), every 8th cm (V8 method), and every 12th cm (V12 method) along the limb. The following formula for a truncated cone was used: $V = \frac{π}{3} h (r_{1}^{2} + r_{2}^{2} + r_{1} \times r_{2})$ 2 where r₁ was the proximal circumference and r₂ was the distal circumference for the segment of interest. The sum of all segments was then determined, giving the full limb volume. The measurement method has shown high reliability and low measurement errors in LLL.¹⁰

To ensure that the same limb length measurement was used for all methods, the length for the V4 method was used as a preference. The most proximal volume segment was therefore converted to either a 4 cm segment or an 8 cm segment for the V8 method or the V12 method (Table 1). To analyze volume measurements in both limbs for all participants, the limb with the larger volume was referred to as the more affected (MA) limb and the limb with the smaller volume was referred to as the less affected (LA) limb in participants with bilateral LE. For participants with unilateral LE, the affected limb was referred to as the MA limb and the nonaffected limb was referred to as the LA limb.

Table 1.

Total Length of the Lower Limbs and Number of Measuring Points for Circumferential Measurements for the V4, V8, and V12 Methods

Total length of the lower limb measured (cm)	V4 method, number of measuring points	V8 method, number of measuring points	V12 method, number of measuring points
70	16	7, with a 4 cm segment as the top cone	6
74	17	9	6, with a 4 cm segment as the top cone
78	18	9, with a 4 cm segment as the top cone	6, with an 8 cm segment as the top cone
82	19	10	7

CMs, circumferential measurements; V4, CMs every 4th cm; V8, CMs every 8th cm; V12, CMs every 12th cm.

Procedure

Measurements of body weight and LLV were performed at two occasions, 2 weeks apart, by an experienced physiotherapist (C.J.). The participants were asked to maintain their normal activity level the day before each test occasion. The measurements took place during the morning about the same time, and with the same procedure each time, starting with the body weight and then 10 minutes of rest in a supine position followed by the CMs.

To identify and mark the measuring points for the CMs, a 110-cm-long measuring board, a 20-cm-long ruler, a narrow measuring tape, and a water-soluble pen were used. To ensure the right position of the LL, the foot and heel were placed against the footplate so that the length measurements on the lateral side of the board were visible (Fig. 1A). The short ruler was used to correctly position the measuring points by reading the length measurements on the board and to mark the points on the limb with 4-cm intervals starting 10 cm above the heel and ending near the groin (Fig. 1B).^8,10

FIG. 1.

(A) The position of the lower limb on the measurement board. (B) Marking the first measurement point 10 cm above the heel. (C) The tape slightly overlapping the measuring point.

For the CMs, the measuring tape was placed around the limb tight to the skin with the tape slightly overlapping the measuring point (Fig. 1C). The measure to the nearest millimeter was taken once at each marking.^8,10 Only the total limb length measure was available at the second test occasion.

To characterize the participants, the following measurements were conducted on the first test occasion: body height and weight, ratings of heaviness and tightness, thickness of the subcutaneous tissue, and leisure time physical activity status.

Statistics

For statistical analysis, IBM (Armonk, NY) SPSS Statistics version 24 was used. Demographics and clinical characteristics of the participants are presented as frequencies, means, and standard deviations (SDs). LLV measurements for each of the three methods on both test occasions are presented as means and SDs.

For the agreement analysis, data from the first test occasion were used. The agreement between the V4 and V8 methods and between the V4 and V12 methods was analyzed by intraclass correlation coefficient (ICC_3.1), and by quantifying the differences between the methods using Bland–Altman graphs. In the graphs, the difference between two methods was plotted against the mean of the two methods to visually demonstrate any systematic bias and outliers.¹⁴ The 95% limits of agreement (LOA) were also calculated where 95% of the differences between the measurements by the two methods are expected to lie.¹⁴

The intrarater test–retest reliability was analyzed by ICC_2.1. According to Fleiss,²¹ ICC values ≤0.40 represent poor reliability and ≥0.75 represent excellent reliability. Changes in the mean were analyzed by calculating mean differences ( $\bar{d}$ ) between the two test occasions (test 2 − test 1) including the 95% confidence interval (95% CI) for $\bar{d}$ to detect any systematic differences between the two test occasions. No systematic change in the mean is present if zero is included in the 95% CI.¹⁴ The SEM and the smallest real difference (SRD) were calculated to determine measurement errors. The SEM gives the limit for the smallest change that indicates a real change for a group of persons²² and is defined as follows: $S E M = S D {(1 - I C C)}^{0.5}$ .²³

The SRD represents the limit for the smallest change that indicates a real change for a single person and is defined as follows: SRD = 1.96 × SEM × $\sqrt{2}$ .²² To make the results easier to interpret, the relative terms (SEM% and SRD%) were also calculated as follows: SEM% = (SEM/mean) × 100²² and SRD% = (SRD/mean) × 100.²⁴ For a group of persons, an acceptable measurement variability (SEM%) is considered to be <10% and for a single individual (SRD%) <30%.²⁴

Results

Participants

Characteristics of the 42 participants (30 women and 12 men) are presented in Table 2. Thirty of them (71%) had secondary cancer-related LLL. Unilateral LE was present in 24 participants (57%), while 18 (43%) had bilateral LE. The duration of the LE was on average 11 (SD 8) years. In the MA limb, thickness of the subcutaneous tissue was present in the lower leg (n = 35) and in the thigh (n = 33). A feeling of heaviness during the past week was reported in 18 participants (43%), while 8 (19%) reported a feeling of tightness. Half of the participants (n = 22) were working, and the level of physical activity varied widely.

Table 2.

Characteristics of the 42 Participants with Lower Limb Lymphedema

Characteristics
Sex, women/men, n (%)	30 (71)/12 (29)
Age, years, mean (SD)	61 (14)
Body mass index (kg/m²), mean (SD)	27 (5)
Primary/secondary LE, n (%)	12 (29)/30 (71)
Cancer diagnosis in secondary LE, n (%)
Gynecological/melanoma/urological/other	17 (40)/5 (12)/4 (10)/4 (10)
Duration of LE, years, mean (SD)	11 (8)
Bilateral/unilateral LE, n (%)	18 (43)/24 (57)
Volume difference in unilateral LE, n (%)
<5%/≥5% to <10%/≥10% to <20%/≥20% to <30%	8 (33)/6 (25)/7 (29)/3 (13)
Location palpated thickness, n (%)
Lower leg, MA limb	35 (83)
Lower leg, LA limb	3 (7)
Thigh, MA limb	33 (79)
Thigh, LA limb	3 (7)
Experience of heaviness, n (%)/median (min–max)
MA limb	18 (43)/35 (13–75)
LA limb	3 (7)/40 (18–53)
Experience of tightness, n (%)/median (min–max)
MA limb	8 (19)/43 (17–67)
LA limb	1 (2)/67
Graded classification system for physical exercise
Score (1–6), median (min–max)	4 (2–6)
Working, n (%)/retired, n (%)	22 (52)/20 (48)

LA, less affected; LE, lymphedema; MA, more affected; SD, standard deviation.

In Table 3, the mean values (SD) of LLV measurements for the V4, the V8, and the V12 methods are presented. On average, there were 14 days (SD 2) between the two test occasions.

Table 3.

Lower Limb Volume Measurements on Two Test Occasions in 42 Participants with Lower Limb Lymphedema Using the V4, V8, and V12 Methods

	V4 method Mean (SD)	V8 method Mean (SD)	V12 method Mean (SD)
Test occasion 1
MA limb, mL	9366 (1548)	9396 (1554)	9401 (1549)
LA limb, mL	8680 (1531)	8708 (1543)	8732 (1504)
Test occasion 2
MA limb, mL	9368 (1552)	9389 (1558)	9407 (1556)
LA limb, mL	8702 (1565)	8729 (1581)	8759 (1531)

Agreement between the measurement methods

In Table 4, agreement between the three measurement methods is presented. For the V4 and V8 methods, ICC was 0.999 and the 95% CI was narrow. The mean difference for the MA limb was −31 mL (95% CI −43 to −18) and for the LA limb −28 mL (95% CI −42 to −13).

Table 4.

Agreement Between Measurements for the V4 and V8 Methods, and Between the V4 and V12 Methods, in the More Affected Limb and Less Affected Limb, Respectively, in 42 Persons with Lower Limb Lymphedema

	ICC_3.1	95% CI for ICC	$\bar{d}$	95% CI of $\bar{d}$	95% LOA
V4 and V8 methods
MA limb	0.999	0.998 to 1.000	−31	−43 to −18	−110 to 49
LA limb	0.999	0.998 to 1.000	−28	−42 to −13	−117 to 62
V4 and V12 methods
MA limb	0.998	0.996 to 0.999	−35	−61 to −9	−198 to 129
LA limb	0.998	0.994 to 0.999	−52	−81 to −23	−236 to 132

CI, confidence interval; $\bar{d}$ , mean difference between the methods; ICC, intraclass correlation coefficient; LOA, limits of agreement.

For the V4 and V12 methods, ICC was 0.998 for the MA and LA limbs, respectively, and the 95% CIs were narrow. The mean difference for the MA limb was −35 mL (95% CI −61 to −9) and for the LA limb −52 mL (95% CI −81 to −23).

The Bland–Altman graphs (Fig. 2A–D) revealed that the variability between the V4 and V8 methods (Fig. 2A, B) and the V4 and V12 methods (Fig. 2C, D) was small. No systematic relationship between the differences was revealed, or no increase in variability for larger volumes was disclosed. For the V4 and V8 methods, the 95% LOA ranged between −117 and 62 mL for the MA and LA limbs, respectively. For the V4 and V12 methods, the 95% LOA ranged between −236 and 132 mL for the MA and LA limbs, respectively.

FIG. 2.

Bland–Altman graphs where the differences between the V4 and V8 methods (A, B) and between the V4 and V12 methods (C, D) were plotted against the mean of the methods, for each limb separately.

Intrarater test–retest reliability

In Table 5, test–retest reliability data are presented for the V4, V8, and V12 methods. The ICCs ranged from 0.993 to 0.995 and the 95% CIs were narrow for all methods. The $\bar{d}$ was small for all three methods in both limbs and no systematic differences were revealed. For all three methods, the SEM% ranged from 1.2% to 1.4% and the SRD% ranged from 3.4% to 3.8% in both limbs.

Table 5.

Intrarater Test–Retest Reliability of Volume Measurements for the V4, V8, and V12 Methods, in the More Affected Limb and Less Affected Limb in 42 Persons with Lower Limb Lymphedema

	ICC_2.1	95% CI for ICC_2.1	$\bar{d}$	95% CI for $\bar{d}$	SEM	SEM%	SRD	SRD%
V4 method
MA limb	0.994	0.989 to 0.997	3	−51 to 56	120	1.3	333	3.6
LA limb	0.995	0.991 to 0.997	21	−26 to 69	108	1.2	299	3.4
V8 method
MA limb	0.994	0.988 to 0.997	−8	−62 to 47	120	1.3	333	3.5
LA limb	0.994	0.990 to 0.997	21	−31 to 72	120	1.4	333	3.8
V12 method
MA limb	0.993	0.987 to 0.996	7	−51 to 64	130	1.4	360	3.8
LA limb	0.994	0.988 to 0.997	26	−27 to 79	117	1.3	324	3.7

$\bar{d}$ , mean difference between test occasion 2 and test occasion 1; SEM, standard error of measurement; SEM%, standard error of measurement in relative terms; SRD, smallest real difference; SRD%, smallest real difference in relative terms.

Discussion

To the best of our knowledge, this is the first study showing that fewer CMs than every 4th cm can be used for LLV measurements in persons with mild-to-moderate LLL without compromising reliability. Overall, the agreement was high between all measurement methods, but slightly higher between the V4 and V8 methods than between the V4 and V12 methods, and the test–retest reliability was equally high for all three methods. The analyses indicate that the V8 method, in particular, can detect clinically relevant changes in LLV measurements similar to the V4 method. The V8 method can thus replace the V4 method when assessing LLV in persons with mild-to-moderate LLL.

The agreement between the V4 and V8 methods and between the V4 and V12 methods was very high (ICC 0.998–0.999), and the differences between the methods were low as visualized by the narrow 95% LOA (−117 to 62 and −236 to 132 mL, respectively). The Bland–Altman graph has been used in previous studies where agreement between the LLV measurements derived from different measurement methods has been evaluated.^25,26 In these studies, a lack of agreement was found in most of the comparisons due to the wide LOA. However, for the comparison of the WDM and the tape measurement method using CMs every 3rd cm in healthy lower legs, the LOA were narrow, indicating that the more practical tape measurement method could be used instead of the WDM.²⁵

Even though the 95% LOA are a matter of clinical consideration, the LOA in our study for the V4 and V8 methods are in line with that of Sukul et al,²⁵ supporting that the agreement between these two methods is acceptable.

Furthermore, the test–retest reliability was very high for all three measurement methods in our study (ICC 0.993–0.995). The findings agree with the results in LLs of healthy persons (ICC 0.99)⁸ and in persons with LLL (ICC 0.99)¹⁰ using CMs every 4th cm for volume measurements. The findings are also in line with the results of persons with breast cancer-related upper limb LE (ICC 0.97–0.99)^27–29 indicating that the tape measurement method is a reliable measurement method to assess volume of the upper and LL.

In our reliability analysis, several statistical methods were used as being recommended.²² The measurement errors (SEM% and SRD%) for the V4, V8, and V12 methods were very low, and in agreement with studies in healthy women and men (SEM% 1.1–1.3%; SRD 3.1–3.6%).⁸ The measurement errors presented in the reliability studies of upper limb volume measurements^27–29 were in absolute values. Relative values would have been preferred as this would have enabled comparisons between studies and facilitated the interpretation for clinical use.²³

For a measuring method to be useful in the clinic, not only reliability but also time efficiency and feasibility must be considered.^29,30 De Vrieze et al²⁹ concluded that the tape measurement method using CMs every 4th cm was the best measuring method considering reliability and measurement error, cost, limitations, and time consumption when different volume measurement methods in upper limb LE were compared. Surprisingly, there are very few studies focusing on these aspects even though the tape measurement method is commonly used in LLL.

In our study, between 16 and 19 measuring points were used for the V4 method, and this is rather time-consuming. The number decreased to 8–10 points for the V8 method and to 6–7 points for the V12 method. Hence, with the V8 method instead of the V4 method, fewer CMs will be used, and the time required for this measurement method becomes shorter. Brorson and Höijer³¹ evaluated upper limb volume measurements based on CMs every 4th cm and CMs at 5 measuring points equivalent to a made-to-measure compression garment. They concluded that volume measurements derived from these two measurement methods did not differ significantly, and by using the same 5 measuring points for volume and for a made-to-measure compression garment, time will be saved.

To evaluate LLV measurements based on CM at measuring points equivalent to a made-to-measure compression garment may also shorten the measurement time, but this measurement method must be investigated using a comprehensive set of statistical analysis such as those used in our study.

Strengths and limitations

A strength of the present study was that 42 participants were included. In reliability studies, a sample size of at least 30 is recommended.^21,32 Another strength was the use of a highly standardized test protocol that had previously been evaluated^8,10 and the comprehensive set of statistical methods for the agreement analysis¹⁴ and the test–retest reliability,^22–24 enabling a comparison between reliability studies in the future. A limitation of the present study was that LLV measurements were based only on CMs on the limb and not on the foot. A single CM on the foot has been shown to be reliable,³³ but to calculate the volume of the foot using the tape measurement method is difficult due to the foot's irregular configuration.

Another limitation is that the result is based on persons with mild-to-moderate LLL. To use the V8 method for LLV measurements in persons with severe LLL needs further investigation as there may be a risk of larger variability in larger volume measurements over time.

Conclusion

The agreement was high between all measurement methods, but slightly higher between the V4 and V8 methods than between the V4 and V12, and the test–retest reliability was equally high for all three methods. The V8 method can thus replace the V4 method when assessing LLV in persons with mild-to-moderate LLL.

Footnotes

Acknowledgments

We thank the persons who volunteered to participate and RPT, PhD, Michael Miller for language editing.

Authors' Contributions

Conceptualization, methodology, formal analysis, and writing—original draft (C.J.). Conceptualization, methodology, and writing—review and editing (K.J., M.B., and C.B.).

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was supported by grants from the Swedish Cancer Society (CAN 2015/443).

References

International Lymphedema Framework 2006 Best practice for the management of lymphedema. International consensus. London. Medical Education Partnership. Available at: www.lympho.org/wp-content/uploads/2021/09/Best_practice.pdf [Last accessed: October 1, 2021).

Brorson

, Svensson

, Ohlin

Chapter 11. Volume measurements and follow-up. In: Lymphedema, Presentation, Diagnosis, and Treatment (Greene AK, Slavin SA, Brorson H. eds.) Springer International Publishing: Cham, Switzerland; 2015; pp. ss115–ss122.

Pasley

, O'Connor

. High day-to-day reliability in lower leg volume measured by water displacement. Eur J Appl Physiol, 2008; 103:393–398; doi: 10.1007/s00421-008-0719-5

Stanton

AWB

, Northfield

, Holroyd

, et al. Validation of an optielectronic limb volumeter (Perometer®). Lymphology, 1997; 30:77–97.

Sharkey

, King

, Kuo

, et al. Measuring limb volume: Accuracy and reliability of tape measurement versus perometer measurements. Lymphat Res Biol, 2018; 16(2):182–186; doi: 10.1089/lrb.2017.0039

Reza

, Norregard

, Moffatt

, et al. Inter-observer and intra-observer variability in volume measurements of the lower extremity using perometer. Lymphat Res Biol, 2020; 18(5):416–421; doi: 10.1089/lrb.2019.0063

Sawan

, Mugnai

, de Barros Lopes

, et al. Lower-limb lymphedema and vulval cancer: Feasibility of prophylactic compression garments and validation of leg volume measurement. Int J Gynecol Cancer, 2009; 19:1649–1654; doi: 10.1111/IGC.0b013e3181a8446a

Jönsson

, Bjurberg

, Brogårdh

, et al. Test-retest reliability of volume and local tissue water measurements in lower limbs of healthy women and men. Lymphat Res Biol, 2020; 18(3):261–269; doi: 10.1089/lrb.2019.0044

Stanton

AWB

, Badger

, Sitzia

. Non-invasive assessment of the lymphedematous limb. Lymphology, 2000; 33:122–135.

10.

Jönsson

, Johansson

, Bjurberg

, et al. Impedance of extracellular fluid, volume and local tissue water can be reliably measured in persons with lower limb lymphedema. Phys Ther, 2022; 102(5):pzac025; doi: 10.1093/ptj/pzac025

11.

Tidhar

, Armer

, Deutscher

, et al. Measurement issues in anthropometric measures of limb volume change in persons at risk for and living with lymphedema: A reliability study. J Pers Med, 2015; 5:341–353; doi: 10.3390/jpm5040341

12.

Mayrovitz

, Macdonald

, Davey

, et al. Measurement decisions for clinical assessment of limb volume changes in patients with bilateral and unilateral edema. Phys Ther, 2007; 87(10):1362–1368; doi: 10.2522/ptj.20060382

13.

Bland

, Altman

. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet, 1986; 1(8476):307–310.

14.

Bland

, Altman

. Measuring agreement in method comparison studies. Stat Methods Med Res, 1999; 135–160; doi: 10.1177/096228029900800204

15.

Scott

, Huskinsson

. Graphic representation of pain. Pain, 1976; 2:175–184.

16.

Johansson

, Lie

, Ekdahl

, et al. A randomized study comparing manual lymph drainage with sequential pneumatic compression for treatment of postoperative arm lymphedema. Lymphology, 1998; 31:56–64.

17.

Johansson

, Albertsson

, Ingvar

, et al. Effects of compression bandaging with or without manual lymph drainage treatment in patients with postoperative arm lymphedema. Lymphology, 1999; 32:103–110.

18.

Jönsson

, Johansson

. The effects of pole walking on arm lymphedema and cardiovascular fitness in women treated for breast cancer: A pilot and feasibility study. Physiother Theory Pract, 2014; 30:236–242; doi: 10.3109/09593985.2013.848961

19.

Brodin

Fysioterapi. Studentlitteratur: Lund, Sweden, 1971.

20.

Frändin

, Grimby

. Assessment of physical activity, fitness and performance in 76 year olds. Scan J Med Sci Sports, 1994; 4:41–46. doi: 10.1111/j.1600-0838.1994.tb00404.x

21.

Fleiss

JL.

The Design and Analysis of Clinical Experiments. John Wiley & Sons: New York, NY, USA; 1986.

22.

Lexell

, Downham

. How to assess the reliability of measurements in rehabilitation. Am J Phys Med Rehabil, 2005; 84:719–723; doi: 10.1097/01.phm.0000176452.17771.20

23.

Downham

, Holmbäck

, Lexell

Chapter 9. Reliability of measurements in medical research and clinical practice. In: Studies in Multidisciplinarity 3 (Paton R, McNamara L, eds.) Elsevier: Amsterdam; 2006; pp. ss147–ss163.

24.

Flansbjer

, Holmbäck

, Downham

, et al. What change in isometric knee muscle strength can be detected in men and women with hemiparesis after stroke?. Clin Rehabil, 2005; 19:514–522; doi: 10.1191/0269215505cr854oa

25.

Sukul

, den Hoed

, Johannes

, et al. Direct and indirect methods for the quantification of leg volume: Comparison between water displacement volumetry, the disk model method and the frustum sign model method, using the correlation coefficient and the limits of agreement. J Biomed Eng, 1993; 15(6):477–480; doi: 10.1016/0141-5425(93)90062-4

26.

Tan

, Coutts

, Bulley

. Measurement of lower limb volume: Agreement between the vertically oriented perometer and a tape measure method. Physiotherapy, 2013; 99(3):247–251; doi: 10.1016/j.physio.2012.12.004

27.

Czerniec

, Ward

, Refshauge

, et al. Assessment of breast cancer-related arm lymphedema—comparison of physical measurement methods and self-report. Cancer Invest, 2010; 28:54–62; doi: 10.3109/07357900902918494

28.

Katz-Leurer

, Bracha

. Test-retest reliability of arm volume measurement in women with breast cancer-related lymphoedema. J Lymphedema, 2012; 2:8–13.

29.

De Vrieze

, Gebruers

, Tjalma

, et al. What is the best method to determine excessive arm volume in patients with breast cancer-related lymphoedema in clinical practice? Reliability, time efficiency and clinical feasibility of five different methods. Clin Rehabil, 2019; 33(7):1221–1232; doi: 10.1177/0269215519835907

30.

Hidding

, Viehoff

, Beurskens

CHG

, et al. Measurement properties of instruments for measuring of lymphedema: Systematic review. Phys Ther, 2016; 96:1965–1981; doi: 10.2522/ptj.20150412

31.

Brorson

, Höijer

. Standardised measurements used to order compression garments can be used to calculate arm volumes to evaluate lymphoedema treatment. J Plast Surg Hand Surg, 2012; 46:410–415; doi: 10.3109/2000656X.2012.714785

32.

Hopkins

WG.

Measures of reliability in sports medicine and in science. Sports Med, 2000; 30:1–15; doi: 10.2165/00007256-200030010-00001

33.

Bakar

, Özdemir ÖC, Sevim

, et al. Intra-observer and inter-observer reliability of leg circumference measurement among six observers: A single blinded randomized trial. J Med Life, 2017; 10(3):176–181.