Skip Navigation


Rheumatology Advance Access originally published online on October 2, 2006
Rheumatology 2007 46(3):533-538; doi:10.1093/rheumatology/kel330
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
46/3/533    most recent
kel330v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (7)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Anderson, M. E.
Right arrow Articles by Herrick, A. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Anderson, M. E.
Right arrow Articles by Herrick, A. L.
Related Collections
Right arrow Systemic Sclerosis
Right arrow Vasculitis
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2006. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

The ‘distal–dorsal difference’: a thermographic parameter by which to differentiate between primary and secondary Raynaud's phenomenon

M. E. Anderson, T. L. Moore1, M. Lunt2 and A. L. Herrick1,2

University of Liverpool Academic Rheumatology Unit, University Hospital Aintree, Liverpool L9 7AL, 1University of Manchester Rheumatic Diseases Centre, Hope Hospital, Salford, M6 8HD and 2Arthritis Research Campaign Epidemiology Unit, University of Manchester, Manchester, Ml3 9PT, UK.

Correspondence to: Ariane Herrick, University of Manchester Rheumatic Diseases Centre, Hope Hospital, Salford M6 8HD, UK. E-mail: ariane.herrick{at}manchester.ac.uk


   Abstract
Top
 Abstract
Introduction
 Patients and methods
 Thermography
 Results
 Discussion
 Acknowledgement
 References
 
Objective. To test the hypothesis that in a patient with Raynaud's phenomenon (RP), a difference of >1°C between the fingertips and the dorsum of the hand [‘distal–dorsal difference’ (DDD), fingers cooler] is specific for underlying structural vascular disease as occurs in systemic sclerosis (SSc), and to evaluate other thermographic parameters in the separation of secondary from primary RP.

Methods. A retrospective analysis of the case notes and thermography results of patients attending thermography, primarily over a 2-yr period. Multinomial logistic regression was used to ascertain whether thermography variables differed between groups with primary RP (56 patients), undifferentiated connective tissue disease (21 patients) and SSc (45 patients), with adjustment for age, sex and smoking.

Results. A DDD >1°C in any finger at 30°C had a positive predictive value of 70%, and a negative predictive value of 82%, in identifying the patient with RP secondary to SSc. From the results of the multinomial logistic regression, a score was derived incorporating age, number of fingers with DDD >1°C at 30°C and maximum rewarming gradient. This score (with a suitable cut-off) was 82% sensitive and 82% specific in identifying RP secondary to SSc, with a positive predictive value of 73% and a negative predictive value of 89%.

Conclusion. Parameters derived from thermography (incorporating both a heat and cold challenge) aid in the prediction of SSc in patients with RP.

KEY WORDS: Thermography, Raynaud's phenomenon, Systemic sclerosis


   Introduction
Top
 Abstract
 Introduction
Patients and methods
 Thermography
 Results
 Discussion
 Acknowledgement
 References
 
Raynaud's phenomenon (RP), episodic colour change of the digits usually in response to cold, is common and results in significant morbidity in the general population. Although RP is usually idiopathic [primary Raynaud's phenomenon (PRP)], when it does not result in tissue damage, RP can occur as part of an underlying disorder; in systemic sclerosis (SSc), RP is associated with characteristic abnormalities in the function and morphology of the vasculature which can result in irreversible digital ischaemia. The vasospasm of PRP and SSc each follow a different clinical course. It is therefore important to possess reliable, objective measures of blood flow to apply in a patient presenting with RP, allowing

  1. the clinician to make the correct diagnosis and instigate appropriate management of the patient, and
  2. the researcher to devise studies which will further our insight into the pathophysiology of RP, and objectively assess new therapies.
These methods should have a high specificity for discriminating between the patient with PRP and RP secondary to SSc. They should be sufficiently sensitive to allow quantification of disease severity and assessment of disease progression over time and in response to treatments.

Thermography has been used for assessment of the patient with RP in specialist centres for over 25 yrs. An infrared camera measures the temperature distribution of the skin by detecting its natural thermal radiation, and thus gives an indirect record of micro- and macro-vascular flow, as the blood carries heat to the imaged area [1]. However, the technique has limitations, including concerns about reproducibility when assessing response to cold challenge [2–4].

In a pilot study, we suggested that the presence of a difference in temperature between one or more of the fingertips and the dorsum of the hand of >1°C (fingers cooler than dorsum) at a room temperature of 30°C is specific for underlying structural vascular disease [5]. To test this hypothesis and further evaluate the different parameters measured during our standard thermographic protocol, we carried out a retrospective analysis of the case notes and thermography results of patients who had attended our vascular laboratory for thermographic testing over a 2-yr period.


   Patients and methods
Top
 Abstract
 Introduction
 Patients and methods
Thermography
 Results
 Discussion
 Acknowledgement
 References
 
Patients
All 161 patients who had attended the vascular laboratory at Hope Hospital for thermographic testing during a 2-yr period were initially included in the study. Of these 161 patients, case notes were available on 152 patients. These case notes were reviewed by the same clinician (M.E.A., a clinical research fellow with an interest in scleroderma-spectrum disorders). Patients were ascribed the appropriate diagnostic categories, based on clinical and serological evidence, without knowledge of the thermography results. Of the 152 patient case notes reviewed:

  1. fifty-six patients had PRP, as classified by LeRoy and Medsger [6],
  2. thirty-one patients had SSc with RP (see subsequent text),
  3. twenty-one patients were classified as having undifferentiated connective tissue disease (UCTD) with RP. None of these subjects had any evidence of sclerodermatous skin involvement nor digital pitting, but did have features precluding a diagnosis of PRP. All patients had full immunological testing [including ANA (positivity defined as IgG ANA titre of ≥1/100), antibodies to extractable nuclear antigens (ENA) and anticentromere antibodies (ACA)], other than one patient in whom ACA status was not available. Four patients in this group had abnormal nailfold capillaries (all UCTD patients had nailfold capillary microscopy),
  4. ten patients had RP secondary to a condition other than (ii) or (iii) (two with hand-arm vibration syndrome, one Jo-1 antibody disease, three rheumatoid arthritis, one primary inflammatory arthritis, one mixed connective tissue disease, one primary Sjogren's syndrome and one secondary to neoplasia), and
  5. thirty-four patients were unclassifiable (either insufficient clinical information or no convincing history of RP on retrospective case note review).

All patients with SSc, who had been added to the Hope Hospital SSc database of patients and who had undergone thermographic testing during the 12 months following the 2-yr study period (14 patients) were added to the analysis to expand the number of SSc patients to 45. In this group of 45 SSc patients, 36 fulfilled the American College of Rheumatology (formerly the American Rheumatism Association) criteria for disease [7]. Nine SSc patients did not fulfil the disease criteria: all nine of these patients had definite sclerodactyly, and eight out of the nine patients had positive autoantibodies (seven ANA and one ACA positive) and/or abnormal nailfold capillary microscopy (six patients). Of the 45 patients with SSc, 36 had limited cutaneous SSc (LCSSc) and 9 had diffuse cutaneous SSc (DCSSc) as defined by LeRoy et al. [8].

The demographic data on the 122 PRP, SSc and UCTD patients included in the analysis are outlined in Table 1. Groups 4 and 5 were excluded from subsequent analysis.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Demographic details of PRP, SSc (LCSSc and DCSSc) and UCTD patients included in statistical analysis of thermographic parameters

 

   Thermography
Top
 Abstract
 Introduction
 Patients and methods
 Thermography
Results
 Discussion
 Acknowledgement
 References
 
All measurements and analyses of thermal images were conducted by a senior vascular technician.

(A) Thermography measurement protocol
All patients were asked not to smoke or consume caffeine-containing beverages for at least 4 h prior to the test. Thermographic images were acquired using an Inframetrics 600M infrared thermography camera for the first 8 months of the study period and following this, using an Agema 570 infrared thermography camera (Flir systems Ltd).

After initial acclimatization of each patient at 23°C for 20 min in a climate-controlled room, thermographic images were taken of the dorsal aspect of both hands. The patient then put on thin latex gloves and placed his/her hands into 15°C water for 1 min. Immediately following this, the gloves were removed and thermographic data of the dorsal aspect of both hands were recorded for 15 min during rewarming. If a difference between the temperature of one or more of the fingertips and the dorsum of the hand of >1°C (fingers cooler than dorsum) existed at a room temperature of 23°C, then the room temperature was raised to 30°C and the patient acclimatized for a further 20 min at this higher temperature before the dorsal aspect of the hands was re-imaged.

(B) Thermography image analysis
Thermographic data was handled by the following software programs on a standard personal computer:

  1. the commercially available Thermoteknix TTX program (version 2.5), for videotaped data obtained with the Inframetrics 600M infrared thermography camera, and
  2. the commercially available Agema Research 2.1 program (with modifications allowing real-time temperature measurement from four fingers, calculation of maximum rewarming gradient and calculation of lag time), for digital data obtained with the Agema 570 infrared thermography camera.
(a) The ‘distal–dorsal difference’

For the images equilibrated at 23°C and 30°C, the following were measured.

  1. Mean temperature for the dorsum of each hand, and
  2. mean temperature within a box defined between the nailbed and the distal interphalangeal joint for each of the index, middle, ring and little fingers.

From these temperature measurements ‘distal–dorsal differences’ (DDDs) were calculated. The DDD for each measured finger was calculated by subtracting the fingertip temperature from the dorsum temperature. Therefore, if the finger was colder than the dorsum, the DDD was positive. For each patient, at both 23 and 30°C room temperatures, the presence or absence of a DDD >1°C in any finger was recorded. In addition, the following DDD variables were derived for each patient at room temperatures of both 23 and 30°C:

  1. maximum DDD across all fingers of both hands and
  2. number of fingers with a DDD >1°C. In patients whose fingertips were already well perfused at a room temperature of 23°C (no fingers with a DDD >1°C at 23°C) and who were therefore not tested at 30°C, the number of fingers with a DDD >1°C was also taken to be zero at 30°C room temperature, and the maximum DDD at 30°C was taken to be equivalent to that at 23°C.
(b) The ‘rewarming curve’ after cold challenge

During the 15 min rewarming period, the temperature of the area between the nailbed and the distal interphalangeal joint was measured continuously for the index, middle and ring fingers of both hands. These data were handled by a software that derives a curve for the 15 min rewarming. The following variables were obtained from the rewarming curve using the custom-written software.

  1. Lag time: the lag phase between the end of the cold challenge test and the onset of rewarming. The minimum lag time that can be recorded using this software is 0.5 min. In patients who have not rewarmed at all at 15 min, the lag time was taken to be 15 min,
  2. maximum temperature recovery rate: the maximum gradient of the rewarming curve, and
  3. percentage recovery: the temperature rise/initial temperature drop x 100% at 2, 5, 10 and 15 min.

These values were averaged across both hands giving a single value for each patient (as in [5]).

Statistical methods
All statistical analyses were conducted using SPSS for windows (version 10.0) on a standard personal computer.

Handling of missing values
All thermographic parameters, as described in (B), were included in the statistical analysis. As outlined in (B) (a), for patients whose fingertips were already well perfused at a room temperature of 23°C (no fingers with a DDD >1°C at 23°C) and were therefore not tested at 30°C (26 of the 56 PRP patients, 8 of the 21 UCTD patients and 5 of the 45 SSc patients), the number of fingers with a DDD >1°C was also taken to be zero at 30°C room temperature, and the maximum DDD across all fingers of both hands at 30°C was recorded as the same as that for 23°C. Other than one SSc patient (in whom it was only possible to analyse five out of eight fingers), all eight fingers were available for analysis for each patient. Cold challenge results were missing for two UCTD patients and three SSc patients—this part of the test was not possible in these five patients for clinical reasons (i.e. severe concurrent digital ischaemia ± digital ulceration precluding cold challenge).

Numbers in the DCSSc subgroup were too small to allow for meaningful comparisons with the other groups, therefore, the DCSSc and LCSSc patient groups were combined in statistical analysis.

Logistic regression
Multinomial logistic regression was used to ascertain whether variables differed between the PRP, UCTD and SSc groups, with adjustment for age, sex and smoking.

All of the thermographic variables plus age (as age, but no other demographic factor, affected the outcome of the original multinomial logistic regression) were then included in a single multinomial logistic regression model, using forward selection of variables (i.e. the single most significant predictor was added to the model, then the next most significant, until no variables remained that improved the model significantly). The model was then simplified by constraining parameters of the model (for example, assuming that the difference between SSc and PRP was the same as the difference between SSc and UCTD), if this did not have a significant adverse effect on the fit of the model.

From the latter analysis, the significant variables were used to compute the predicted probability of being in each of the three groups, given the thermographic variables, using logistic regression. Logistic regression calculates the log of the odds of a given outcome as a linear function of the predictor variables, so the prediction equations took the form


Formula

where p is the probability of belonging to the group, x1 and x2 are the variables in the prediction model and {alpha} and ß are regression coefficients.

From this analysis a simplified score was calculated, by rounding of the ß coefficients, which discriminated between SSc and the other two groups, and the performance of this score evaluated by measuring its sensitivity, specificity and positive and negative predictive value at a number of thresholds.


   Results
Top
 Abstract
 Introduction
 Patients and methods
 Thermography
 Results
Discussion
 Acknowledgement
 References
 
Table 2 shows PRP, SSc and UCTD group values for each of the DDD and rewarming curve parameters and results of multinomial logistic regression of these individual parameters between groups. All individual DDD and rewarming curve parameters, other than the percentage maximum rewarming recovery achieved, were significantly different for the PRP and SSc groups, whether or not we adjusted for age, sex and smoking. A number of variables differed between the SSc and UCTD groups, but none differed between the PRP and UCTD groups. This suggested that the thermographic variables could not be used to discriminate between PRP and UCTD, but could distinguish both of these groups from SSc.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Distal–dorsal difference (DDD) and rewarming curve variables along with results of multinomial logistic regression comparing PRP, SSc and UCTD patient groups

 
The results for multivariate multinomial logistic regressions for comparing the SSc with the other two groups are given in Table 3. The variables that entered the model were age, number of fingers with DDD >1°C at 30°C and maximum rewarming gradient (in that order). Constraining all the parameters for the PRP to SSc comparison to be equal to the parameters for the UCTD to SSc comparison did not reduce the fit of the model at all, so a logistic regression equation comparing SSc with the pooled PRP and UCTD groups would provide as much information as the multinomial regression.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Summary of multivariate multinomial regression models

 
The logistic regression equation was


Formula

From this equation to predict SSc, a simplified score was derived by doubling the coefficients and rounding off to 1 significant figure:


Formula

This score had a Spearman correlation coefficient >0.95 with the predicted probability of having SSc from the logistic regression equation, showing that little was lost through the simplification. It differed greatly between SSc and the other two groups (Figure 1), and was an extremely good predictor of SSc [area under the receiver operating characteristic (ROC) curve (AUC) = 0.89; Figure 2]. It is also clear from this figure that this score is a significantly better predictor than age alone (AUC = 0.81) or a combination of age, gender and smoking (AUC = 0.83).


Figure 1
View larger version (8K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 1. Boxplots of SSc prediction score for PRP, SSc and UCTD patient groups.

 

Figure 2
View larger version (24K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 2. ROC curve for use of SSc prediction score for discriminating between SSc and other patients.

 
The validity (sensitivity, specificity and positive and negative predictive values) of the score as a stand-alone measure in the correct diagnosis of the patient with RP (PRP vs RP secondary to SSc) was assessed by cross-tabulation of the PRP and SSc patient groups’ results, having made suitable cut-off points for the predictor. These are presented in Table 4 along with validity indices of the presence of a DDD >1°C in any finger at 30°C alone.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Sensitivity, specificity and positive and negative predictive values, in ability to identify the patient with RP secondary to SSc
  • Presence of a DDD > 1°C in any finger at 30°C (DDD 30°C)
  • SSc prediction score (Age/5 + 0.5 * (Number of fingers with DDD > 1°C at 30°C) – Maximum rewarming gradient).

 

   Discussion
Top
 Abstract
 Introduction
 Patients and methods
 Thermography
 Results
 Discussion
Acknowledgement
 References
 
Thermal imaging has been used as a non-invasive investigative tool in medicine since the 1960s, and has obvious attractions for the assessment of the patient with RP. However, technical limitations of first generation infrared cameras restricted the use of thermography in medicine until very recently, with improvements in camera technology and data handling [1, 9]. For example, the Agema camera used in the later part of the period under review is smaller and more portable than our earlier equipment, and does not require liquid nitrogen as a coolant. However, the imaging protocol was exactly the same with both cameras and so the change of camera during the period of this retrospective review would not have influenced our results.

Cold provocation thermography has been shown to have good sensitivity, specificity and good positive and negative predictive values for the diagnosis of digital vasospasm in patients with hand-arm vibration syndrome (vibration white finger) [10]. However, it must be noted that these patients were compared with normal control subjects, and not patients with PRP. Similarly, computerized digital thermometry with cold challenge has been reported as having high sensitivity and specificity for RP in comparison with healthy control subjects [11]. When Schuhfried et al. [12] examined thermographic parameters in the diagnosis of secondary RP, they did not address the issue of which thermographic variables discriminated between primary and secondary RP. Yet the important question for the rheumatologist, with an interest in RP and connective tissue disease, is whether thermography can distinguish between a patient with benign PRP and a patient with RP secondary to another disorder. Our study addressed this issue.

Although our study was retrospective, and hence has the attendant problems associated with this study design, including the need to exclude many potential subjects who had attended for thermography within the studied time period, it allowed us to examine thermograms from a large cohort of patients with established diagnoses. We corroborated the results of our pilot study [5], finding that the presence of a DDD >1°C in any finger at a room temperature of 30°C is specific (83%) for SSc in a population of patients presenting to our rheumatology department with RP. However, we found this parameter was not particularly sensitive (69%) in identifying the patient with SSc. On multinomial logistic regression, there were also significant differences between the PRP and SSc groups for eight of the remaining nine parameters generated from the thermography examination (Table 2). It was also apparent that the age of the patient being referred for thermographic testing was a significant determinant in its own right of whether the patient had PRP or RP secondary to SSc. Therefore, we pursued the best composite variable of these parameters to ascertain if it would be possible to improve upon the sensitivity of the DDD at 30°C, whilst maintaining a good specificity.

The ideal test discriminates perfectly between those with and without disease. However, this is only possible when test results for the disease positive and disease negative groups do not overlap—a rare occurrence in human biology. Sensitivity and specificity are inversely related for any continuous outcome measure, as is illustrated by our prediction score in Table 4: the trade-off for good specificity is poor sensitivity, and vice versa [13].

The good specificity of a DDD >1°C in any finger at a room temperature of 30°C in predicting the patient with RP secondary to SSc, is balanced by a poorer sensitivity of this parameter on its own. As the DDD at 30°C is an ‘all or nothing’ variable (i.e. the patient either has a positive or a negative result) rather than a continuous variable, there is no scope for changing a cut-off point to address the balance of sensitivity and specificity. Judging the importance of the balance between sensitivity and specificity is dependent on the implications of the test. In a tertiary referral centre for RP and SSc, we ideally want to correctly identify all those RP patients who have or who will go on to develop SSc, in order that we can institute the correct management. In this scenario, a sensitive test might seem more valuable than a specific test. However, the downside is a larger number of false-positive results, with the resultant patient anxiety, and some may argue that specificity is more important than sensitivity.

It could also be argued that, in the setting of the out-patient department, where the clinician has to interpret test results, it is more useful to know the predictive value of thermographic results. The positive predictive value gives the percentage chance of having SSc if the thermography test is ‘positive’, whilst the negative predictive value gives the percentage chance of having PRP if the thermography test is ‘negative’. The presence of a DDD at 30°C has a good positive predictive value for SSc (70%) and good negative predictive value of 82%.

The prediction score, with appropriate cut-off values, does improve upon the sensitivity of thermographic testing whilst maintaining good specificity, as outlined in Table 4. A score >9 has 82% sensitivity and 82% specificity for SSc. Positive and negative predictive values are also good for this value of the score: both the positive and negative predictive values are better than that of the DDD at 30°C. As exhibited in Table 4 and explained earlier in the article, the prediction score does improve upon test validity. Although it does require computation of a formula, it is not complicated, and the availability of spreadsheet functions renders this procedure simple.

The results from this study cannot be extrapolated directly to all clinical settings. This study examined the validity of thermographic testing in a cohort of patients, referred to a specialist centre with an interest in RP and SSc. In a ‘high-prevalence’ setting such as this, the predictive value of a positive test is high, but if used in low-prevalence settings, even a good test will have poor positive predictive value. Therefore, thermographic testing in RP is likely to remain the preserve of the specialist vascular laboratory. However, this study outlines that, with careful validation of technique, thermographic testing is a useful investigative procedure in the specialist setting. It would be of interest to compare the sensitivity, specificity, positive and negative predictive values of thermography and nailfold capillaroscopy in the evaluation of the patient with RP. These techniques are complementary in that nailfold capillaroscopy measures microvascular structure, whereas thermography measures vascular function (albeit indirectly). It has been suggested that the sensitivity and specificity of nailfold capillaroscopy in the diagnosis of scleroderma-spectrum disorders are in the order of 83–97 and 89–98%, respectively [14]. It seems likely that a composite index incorporating both thermography and nailfold capillaroscopy might further improve our ability to identify underlying connective tissue disease in patients presenting with RP.

None of the individual thermographic parameters nor any combination of them separated the UCTD from the PRP patients. This could imply that none of the UCTD patients studied had significant structural vascular disease. However, as a portion of these patients will go on to develop a definite CTD such as SSc, it would be valuable to follow this group of patients with serial thermographic tests over time in order to ascertain if they go on to develop thermographic results similar to those of the SSc group of patients.

Whilst the experimental work of this article has been dedicated to assessing how useful thermography is at identifying patients with RP secondary to SSc, the question of how useful this test is in monitoring vascular disease remains unanswered. Further prospective study of the evolution of a DDD at 30°C and the prediction scores over time is necessary.

Also, it must be recognized that several factors may affect thermoregulatory responses. First, it is known that there is considerable intra-subject variability [5], in part reflecting a patient's sympathetic tone at the time of the test. This variability can be minimized by incorporating an acclimatization period, and asking patients to abstain from smoking and from caffeine containing beverages as is our usual protocol. However, in the routine clinical setting there are other influencing factors which are difficult to control for, for example the stage of the menstrual cycle [15] and a patient's drug treatment.

In conclusion, the presence of a DDD >1°C in any finger at a room temperature of 30°C is specific for underlying structural vascular disease and, in combination with patient age and a smaller maximum rewarming gradient, may complement other investigations in alerting the clinician to an increased likelihood of underlying connective tissue disease.

Formula


   Acknowledgement
Top
 Abstract
 Introduction
 Patients and methods
 Thermography
 Results
 Discussion
 Acknowledgement
References
 
We are grateful to the Raynaud's and Scleroderma Association for funding of the thermography camera.

The authors have declared no conflicts of interest.


   References
Top
 Abstract
 Introduction
 Patients and methods
 Thermography
 Results
 Discussion
 Acknowledgement
 References
 

  1. Jones BF. (1998) A reappraisal of the use of infrared thermal image analysis in medicine. IEEE Transactions Med Imaging 17:1019–27.
  2. Chucker F, Fowler RC, Motomiya T, Singh B, Hurley W. (1971) Induced temperature transients in Raynaud's disease measured by thermography. A preliminary report. Angiology 22:580–93.[Web of Science][Medline]
  3. Chucker FD, Fowler RC, Hurley CW. (1973) Photoplethysmometry and thermography in Raynaud's disorders. A preliminary report. Angiology 24:612–8.[Web of Science][Medline]
  4. Herrick A, el-Hadidy K, Marsh D, Jayson M. (1994) Abnormal thermoregulatory responses in patients with reflex sympathetic dystrophy syndrome. J Rheumatol 21:1319–24.[Web of Science][Medline]
  5. Clark S, Hollis S, Campbell F, Moore T, Jayson M, Herrick A. (1999) The "distal-dorsal difference" as a possible predictor of secondary Raynaud's phenomenon. J Rheumatol 26:1125–8.[Web of Science][Medline]
  6. LeRoy EC and Medsger TA. (1992) Raynaud's phenomenon: a proposal for classification. Clin Exp Rheumatol 10:485–8.[Web of Science][Medline]
  7. Masi AT, Rodnan GP, Medsger TA. (1980) Subcommittee for Scleroderma Criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Preliminary criteria for the classification of systemic sclerosis (scleroderma). Arthritis Rheum 23:581–90.[Web of Science][Medline]
  8. LeRoy EC, Black C, Fleischmajer R, et al. (1988) Scleroderma (systemic sclerosis): classification, subsets and pathogenesis. Raynaud's phenomenon: a proposal for classification. J Rheumatol 15:202–5.[Web of Science][Medline]
  9. Anbar M. (1998) Clinical thermal imaging today. IEEE Eng Med Biology Magazine 17:25–33.
  10. Coughlin PA, Chetter IC, Kent PJ, Kester RC. (2001) The analysis of sensitivity, specificity, positive predictive value and negative predictive value of cold provocation thermography in the objective diagnosis of the hand-arm vibration syndrome. Occup Med 51:75–80.[Abstract]
  11. Caramaschi P, Codella O, Poli G, et al. (1989) Use of computerized digital thermometry for diagnosis of Raynaud's phenomenon. Angiology 40:863–71.[Web of Science][Medline]
  12. Schuhfried O, Vacariu G, Lang T, Korpan M, Kiener HP, Fialka-Moser V. (2000) Thermographic parameters in the diagnosis of secondary Raynaud's phenomenon. Arch Phys Med Rehabil 81:495–9.[CrossRef][Web of Science][Medline]
  13. Grimes DA and Schulz KF. (2002) Uses and abuses of screening tests. Lancet 359:881–4.[CrossRef][Web of Science][Medline]
  14. Carpentier PH and Maricq HR. (1990) Microvasculature in systemic sclerosis. Rheum Dis Clin N America 16:75–91.
  15. Lafferty K, de Trafford JC, Potter C, Roberts VC, Cotton LT. (1985) Reflex vascular responses in the finger to contralateral thermal stimuli during the normal menstrual cycle: a hormonal basis to Raynaud's phenomenon. Clin Sci 68:639–45.
Submitted 3 August 2006; revised version accepted 15 August 2006.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Rheumatology (Oxford)Home page
K. A. Binymin
Comment on: Sparing of the thumb in Raynaud's phenomenon
Rheumatology, August 1, 2008; 47(8): 1260 - 1260.
[Full Text] [PDF]

Home page
 Rheumatology (Oxford)Home page
B. Chikura, T. L. Moore, J. B. Manning, A. Vail, and A. L. Herrick
Sparing of the thumb in Raynaud's phenomenon
Rheumatology, February 1, 2008; 47(2): 219 - 221.
[Abstract] [Full Text] [PDF]

Home page
 Rheumatology (Oxford)Home page
H. Gunawardena, N. D. Harris, C. Carmichael, and N. J. McHugh
Microvascular responses following digital thermal hyperaemia and iontophoresis measured by laser Doppler imaging in idiopathic inflammatory myopathy
Rheumatology, September 1, 2007; 46(9): 1483 - 1486.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
46/3/533    most recent
kel330v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (7)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Anderson, M. E.
Right arrow Articles by Herrick, A. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Anderson, M. E.
Right arrow Articles by Herrick, A. L.
Related Collections
Right arrow Systemic Sclerosis
Right arrow Vasculitis
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?