For your review: Chapter 13: Reading massage research--The Abstract

This is a chapter from an upcoming book on massage research literacy, posted for your feedback and review.

The massage research literacy book will be available here on a free and open-access basis, and its expected publication date is 1 September 2012.

Learning objectives for this chapter

Upon completion of this chapter, you will be able to understand, do, and value the following:

Do

• Explain what a research article abstract is (Bloom's learning taxonomy level: explanation)
• Describe how the abstract relates to the IMRaD structure (Bloom's learning taxonomy level: explanation)

Know

• What information about the article that you can expect the abstract to provide (Bloom's learning taxonomy level: explanation)
• Why not to use an abstract as a substitute for reading the article itself (Bloom's learning taxonomy level: explanation)

Appreciate

• What a structured abstract is, and how it helps communicate information to the reader (Bloom's learning taxonomy level: explanation)
• How to use an abstract to evaluate whether an article is worth obtaining (Bloom's learning taxonomy level: explanation)

The final research article section we’ll examine here is usually (although not always) actually the first section in the printed article itself. It’s not included by name in the IMRaD structure, because it does not provide any new information which the other sections don’t. It is simply a brief overview of the entire article itself, so it contains information from all the IMRaD sections in a very concise form.

The abstract as a summary of the entire research article

You can think of the abstract as kind of an executive summary (and indeed, it is sometimes labeled "Summary", rather than "Abstract", sometimes it it not labeled at all, as in this illustration).

It is a small snapshot of the article, intended to give you just enough information to decide whether it is useful enough to you to go to the effort of obtaining and actually reading the entire article.

The author’s job in writing it, then, is to walk a tightrope between two opposite values—the abstract needs to contain enough well-structured, organized information to help you decide if the article is useful to you, yet in presenting that much information, it also needs to be brief, clear, and easy to read. You should come away from reading the abstract knowing the following things about the article:

• What research question was being investigated, and why;
• What population the treatment was studied in;
• What methodology was used to study the treatment;
• What results the investigator got for the outcome, including the statistical significance of those results;
• The conclusions the investigator drew about the study, including the meaning for clinical practice, and the basis for further research.

Structured and unstructured/narrative abstracts

Sometimes the abstract literally follows the IMR(a)D structure of the article very closely, although they may vary the wording somewhat, such as "Objectives" for "Introduction", or "Study Design" for "Methodology", for example.

This style of abstract is called, not surprisingly, a "structured abstract", and is becoming increasingly common in the literature because reseach has demonstrated that it is not only more user-friendly, but it also helps the authors avoid falling into common writing errors.

In the same way that the standard structure of the article itself helps the reader to navigate the information, the structured abstract also helps the reader to efficiently pick out the important points.

In the examples below, I have bolded the sections, and broken the paragraphs where necessary in such a way as to emphasize the parts I especially want you to notice.

Many structured abstracts do that formatting themselves, as in this example from Aly 2004, where the journal makes the different sections of the structured abstract very clear:

Aly H, Moustafa MF, Hassanein SM, Massaro AN, Amer HA, Patel K. Physical activity combined with massage improves bone mineralization in premature infants: a randomized trial. J Perinatol. 2004 May;24(5):305-9. PMID: 15071483

BACKGROUND: Osteopenia of prematurity is a known source for morbidity in preterm infants. Premature infants have shown favorable outcomes in response to massage and physical activity. Whether such intervention can stimulate bone formation or decrease bone resorption is yet to be determined.

OBJECTIVE: To test the hypothesis that massage combined with physical activity can stimulate bone formation and ameliorate bone resorption in premature infants.

DESIGN/METHODS: A prospective double-blinded randomized trial was conducted at the Neonatal Intensive Care Unit of Ain Shams University in Cairo, Egypt. Thirty preterm infants (28 to 35 weeks' gestation) were randomly assigned to either control group (Group I, n=15) or intervention group (Group II, n=15). Infants in the intervention group received a daily protocol of combined massage and physical activity. Serum type I collagen C-terminal propeptide (PICP) and urinary pyridinoline crosslinks of collagen (Pyd) were used as indices for bone formation and resorption, respectively. PICP and Pyd were measured at enrollment and at discharge for all subjects. t-Test, ANOVA and linear regression analysis were used for statistical analyses.

RESULTS: There was no difference between groups I and II in gestational age (32.1+/-1.8 vs 31.5+/-1.4 weeks) or birth weight (1.429+/-0.148 vs 1.467+/-0.132 g). In the control group, serum PICP decreased over time from 82.3+/-8.5 to 68.78+/-14.6 (p<0.01), while urinary Pyd increased from 447.7+/-282.8 to 744.9+/-373.6 (p<0.01) indicating decreased bone formation and increased bone resorption, respectively. In the intervention group, serum PICP increased over time from 62.5+/-13.8 to 73.84+/-12.9 (p<0.01). Urinary Pyd also increased over time from 445.7+/-266.5 to 716.8+/-301.8 (p<0.01). In a linear regression model including gestational age and intervention, serum PICP increased significantly in the intervention group (regression coefficient 18.8+/-4.6, p=0.0001) while urinary Pyd did not differ between groups (regression coefficient=5.6+/-114.3, p=0.961).

CONCLUSIONS: A combined massage and physical activity protocol improved bone formation (PICP) but did not affect bone resorption (Pyd). Pyd increased over time in both groups, possibly due to continuous bone resorption and Ca mobilization.

Still, don't be surprised to see some structured abstracts that are just one long run-on paragraph, as well, as in this example from Ferber 2002.

Ferber SG, Laudon M, Kuint J, Weller A, Zisapel N. Massage therapy by mothers enhances the adjustment of circadian rhythms to the nocturnal period in full-term infants. J Dev Behav Pediatr. 2002 Dec;23(6):410-5. PMID: 12476070

The objective of this study was to investigate the effect of massage therapy on phase adjustment of rest-activity and melatonin secretion rhythms to the nocturnal period in full-term infants. Rest-activity cycles of infants (measurement 1, n = 16) were measured by actigraphy before and after 14 days of massage therapy (starting at age 10 [+/-4] d) and subsequently at 6 and 8 weeks of age. 6-Sulphatoxymelatonin excretion was assessed in urine samples at 6, 8, and 12 weeks of age (measurement 2, n = 21). At 8 weeks the controls revealed one peak of activity at approximately 12 midnight (11 p.m.-3 a.m.) and another one at approximately 12 noon (11 a.m.-3 p.m.), whereas in the treated group, a major peak was early in the morning (3 a.m.-7 a.m.) and a secondary peak in the late afternoon (3 p.m.-7 p.m.). At 12 weeks, nocturnal 6-sulphatoxymelatonin excretions were significantly higher in the treated infants (1346.38 +/- 209.40 microg/night vs 823.25 +/- 121.25 microg/night, respectively; <.05). It is concluded that massage therapy by mothers in the perinatal period serves as a strong time cue, enhancing coordination of the developing circadian system with environmental cues.

In the comments to the review version of this chapter, Rosemary Chunco made an awesome suggestion for dealing with unstructured abstracts:

Oh - and just as a comment on abstract writing in general.  I like the IMRaD structure built in, so when I come across a wordy abstract that is just a paragraph or two that has a study involved (for some types of articles, the IMRaD structure doesn't apply), I transpose the wordiness of it on to an IMRaD structure if I want to examine it in any depth.  It helps clarification.

Since this abstract is a little dense to tease information out of, let's try her technique on it.

First, we'll break it up into individual sentences:

1. The objective of this study was to investigate the effect of massage therapy on phase adjustment of rest-activity and melatonin secretion rhythms to the nocturnal period in full-term infants.
2. Rest-activity cycles of infants (measurement 1, n = 16) were measured by actigraphy before and after 14 days of massage therapy (starting at age 10 [+/-4] d) and subsequently at 6 and 8 weeks of age.
3. 6-Sulphatoxymelatonin excretion was assessed in urine samples at 6, 8, and 12 weeks of age (measurement 2, n = 21).
4. At 8 weeks the controls revealed one peak of activity at approximately 12 midnight (11 p.m.-3 a.m.) and another one at approximately 12 noon (11 a.m.-3 p.m.), whereas in the treated group, a major peak was early in the morning (3 a.m.-7 a.m.) and a secondary peak in the late afternoon (3 p.m.-7 p.m.).
5. At 12 weeks, nocturnal 6-sulphatoxymelatonin excretions were significantly higher in the treated infants (1346.38 +/- 209.40 microg/night vs 823.25 +/- 121.25 microg/night, respectively; <.05).
6. It is concluded that massage therapy by mothers in the perinatal period serves as a strong time cue, enhancing coordination of the developing circadian system with environmental cues.

Sentence 1, the objective of the study, clearly is in the I part of IMR(a)D, so we put it under "Introduction", or "Objective", if you prefer.

1. INTRODUCTION: The objective of this study was to investigate the effect of massage therapy on phase adjustment of rest-activity and melatonin secretion rhythms to the nocturnal period in full-term infants.
2. Rest-activity cycles of infants (measurement 1, n = 16) were measured by actigraphy before and after 14 days of massage therapy (starting at age 10 [+/-4] d) and subsequently at 6 and 8 weeks of age.
3. 6-Sulphatoxymelatonin excretion was assessed in urine samples at 6, 8, and 12 weeks of age (measurement 2, n = 21).
4. At 8 weeks the controls revealed one peak of activity at approximately 12 midnight (11 p.m.-3 a.m.) and another one at approximately 12 noon (11 a.m.-3 p.m.), whereas in the treated group, a major peak was early in the morning (3 a.m.-7 a.m.) and a secondary peak in the late afternoon (3 p.m.-7 p.m.).
5. At 12 weeks, nocturnal 6-sulphatoxymelatonin excretions were significantly higher in the treated infants (1346.38 +/- 209.40 microg/night vs 823.25 +/- 121.25 microg/night, respectively; <.05).
6. It is concluded that massage therapy by mothers in the perinatal period serves as a strong time cue, enhancing coordination of the developing circadian system with environmental cues.

In sentence 2, even if we don't know exactly what "actigraphy" is (we can look it up later), it's still clear that "were measured" means that it's about methods, the "M" in IMR(a)D. In sentence 3, "was assessed" tells us the same thing, so both sentences get a "Methods" label.

1. INTRODUCTION: The objective of this study was to investigate the effect of massage therapy on phase adjustment of rest-activity and melatonin secretion rhythms to the nocturnal period in full-term infants.
2. METHODS: Rest-activity cycles of infants (measurement 1, n = 16) were measured by actigraphy before and after 14 days of massage therapy (starting at age 10 [+/-4] d) and subsequently at 6 and 8 weeks of age.
3. METHODS: 6-Sulphatoxymelatonin excretion was assessed in urine samples at 6, 8, and 12 weeks of age (measurement 2, n = 21).
4. At 8 weeks the controls revealed one peak of activity at approximately 12 midnight (11 p.m.-3 a.m.) and another one at approximately 12 noon (11 a.m.-3 p.m.), whereas in the treated group, a major peak was early in the morning (3 a.m.-7 a.m.) and a secondary peak in the late afternoon (3 p.m.-7 p.m.).
5. At 12 weeks, nocturnal 6-sulphatoxymelatonin excretions were significantly higher in the treated infants (1346.38 +/- 209.40 microg/night vs 823.25 +/- 121.25 microg/night, respectively; <.05).
6. It is concluded that massage therapy by mothers in the perinatal period serves as a strong time cue, enhancing coordination of the developing circadian system with environmental cues.

"Revealed" in sentence 4, and "were significantly higher" in sentence 5, indicate that these sentences belong in the "R" of "IMR(a)D": Results:

1. INTRODUCTION: The objective of this study was to investigate the effect of massage therapy on phase adjustment of rest-activity and melatonin secretion rhythms to the nocturnal period in full-term infants.
2. METHODS: Rest-activity cycles of infants (measurement 1, n = 16) were measured by actigraphy before and after 14 days of massage therapy (starting at age 10 [+/-4] d) and subsequently at 6 and 8 weeks of age.
3. METHODS: 6-Sulphatoxymelatonin excretion was assessed in urine samples at 6, 8, and 12 weeks of age (measurement 2, n = 21).
4. RESULTS: At 8 weeks the controls revealed one peak of activity at approximately 12 midnight (11 p.m.-3 a.m.) and another one at approximately 12 noon (11 a.m.-3 p.m.), whereas in the treated group, a major peak was early in the morning (3 a.m.-7 a.m.) and a secondary peak in the late afternoon (3 p.m.-7 p.m.).
5. RESULTS: At 12 weeks, nocturnal 6-sulphatoxymelatonin excretions were significantly higher in the treated infants (1346.38 +/- 209.40 microg/night vs 823.25 +/- 121.25 microg/night, respectively; <.05).
6. It is concluded that massage therapy by mothers in the perinatal period serves as a strong time cue, enhancing coordination of the developing circadian system with environmental cues.

"It is concluded" in sentence 6, means Conclusions, which means Discussion section: the D in IMR(a)D:

1. INTRODUCTION: The objective of this study was to investigate the effect of massage therapy on phase adjustment of rest-activity and melatonin secretion rhythms to the nocturnal period in full-term infants.
2. METHODS: Rest-activity cycles of infants (measurement 1, n = 16) were measured by actigraphy before and after 14 days of massage therapy (starting at age 10 [+/-4] d) and subsequently at 6 and 8 weeks of age.
3. METHODS: 6-Sulphatoxymelatonin excretion was assessed in urine samples at 6, 8, and 12 weeks of age (measurement 2, n = 21).
4. RESULTS: At 8 weeks the controls revealed one peak of activity at approximately 12 midnight (11 p.m.-3 a.m.) and another one at approximately 12 noon (11 a.m.-3 p.m.), whereas in the treated group, a major peak was early in the morning (3 a.m.-7 a.m.) and a secondary peak in the late afternoon (3 p.m.-7 p.m.).
5. RESULTS: At 12 weeks, nocturnal 6-sulphatoxymelatonin excretions were significantly higher in the treated infants (1346.38 +/- 209.40 microg/night vs 823.25 +/- 121.25 microg/night, respectively; <.05).
6. CONCLUSIONS: It is concluded that massage therapy by mothers in the perinatal period serves as a strong time cue, enhancing coordination of the developing circadian system with environmental cues.

Collapsing sentences with similar labels into one label, and removing the numbers, we get:

INTRODUCTION: The objective of this study was to investigate the effect of massage therapy on phase adjustment of rest-activity and melatonin secretion rhythms to the nocturnal period in full-term infants.

METHODS: Rest-activity cycles of infants (measurement 1, n = 16) were measured by actigraphy before and after 14 days of massage therapy (starting at age 10 [+/-4] d) and subsequently at 6 and 8 weeks of age. 6-Sulphatoxymelatonin excretion was assessed in urine samples at 6, 8, and 12 weeks of age (measurement 2, n = 21).

RESULTS: At 8 weeks the controls revealed one peak of activity at approximately 12 midnight (11 p.m.-3 a.m.) and another one at approximately 12 noon (11 a.m.-3 p.m.), whereas in the treated group, a major peak was early in the morning (3 a.m.-7 a.m.) and a secondary peak in the late afternoon (3 p.m.-7 p.m.). At 12 weeks, nocturnal 6-sulphatoxymelatonin excretions were significantly higher in the treated infants (1346.38 +/- 209.40 microg/night vs 823.25 +/- 121.25 microg/night, respectively; <.05).

CONCLUSIONS: It is concluded that massage therapy by mothers in the perinatal period serves as a strong time cue, enhancing coordination of the developing circadian system with environmental cues.

The unstructured abstract is now easier to read, and if there are parts missing, you'll be able to see right away that they're not there. That's an excellent idea for a technique, Rosemary; thank you very much for suggesting it.

You can get another important aspect from the abstract: a sense of what the experience of reading the entire article will be like.

We can already tell from this abstract that the Ferber 2002 article is highly quantitative, and draws on some knolwedge of biochemistry--the understanding of what 6-sulphatoxymelatonin is, and the meaning about the infant's sleep cycle that its measurement indicates.

You can use the "taste" that the abstract provides to decide whether making the effort to read the entire article is worth it to you, or if you prefer instead to continue looking for other articles that address different aspects of the research on your topic.

Let's look at some examples of structured abstracts from the massage research literature, to see what information they contain:

Hopper D, Deacon S, Das S, Jain A, Riddell D, Hall T, Briffa K. Dynamic soft tissue mobilisation increases hamstring flexibility in healthy male subjects. Br J Sports Med. 2005 Sep;39(9):594-8; discussion 598. PMID: 16118294 PMCID: PMC1725327 Free PMC Article

OBJECTIVES: The purpose of this study was to investigate the effect of dynamic soft tissue mobilisation (STM) on hamstring flexibility in healthy male subjects.

METHODS: Forty five males volunteered to participate in a randomised, controlled single blind design study. Volunteers were randomised to either control, classic STM, or dynamic STM intervention. The control group was positioned prone for 5 min. The classic STM group received standard STM techniques performed in a neutral prone position for 5 min. The dynamic STM group received all elements of classic STM followed by distal to proximal longitudinal strokes performed during passive, active, and eccentric loading of the hamstring. Only specific areas of tissue tightness were treated during the dynamic phase. Hamstring flexibility was quantified as hip flexion angle (HFA) which was the difference between the total range of straight leg raise and the range of pelvic rotation. Pre- and post-testing was conducted for the subjects in each group. A one-way ANCOVA followed by pairwise post-hoc comparisons was used to determine whether change in HFA differed between groups. The alpha level was set at 0.05.

RESULTS: Increase in hamstring flexibility was significantly greater in the dynamic STM group than either the control or classic STM groups with mean (standard deviation) increase in degrees in the HFA measures of 4.7 (4.8), -0.04 (4.8), and 1.3 (3.8), respectively.

CONCLUSIONS: Dynamic soft tissue mobilisation (STM) significantly increased hamstring flexibility in healthy male subjects.

Exercise 1

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Diego MA, Field T, Hernandez-Reif M. Vagal activity, gastric motility, and weight gain in massaged preterm neonates. J Pediatr. 2005 Jul;147(1):50-5. PMID: 16027695

OBJECTIVE: Multiple studies have documented an increase in weight gain after 5 to 10 days of massage therapy for preterm neonates. The massaged preterm neonates did not consume more calories than the control neonates. One potential mechanism for these effects might involve massage-induced increases in vagal activity, which in turn may lead to increased gastric motility and thereby weight gain.

STUDY DESIGN: The present randomized study explored this potential underlying mechanism by assessing gastric motility and sympathetic and parasympathetic nervous system activity in response to massage therapy (moderate pressure) versus sham massage (light pressure) and control conditions in a group of preterm neonates.

RESULTS: Compared with preterm neonates receiving sham massage, preterm neonates receiving massage therapy exhibited greater weight gain and increased vagal tone and gastric motility during and immediately after treatment. Gastric motility and vagal tone during massage therapy were significantly related to weight gain.

CONCLUSION: The weight gain experienced by preterm neonates receiving moderate-pressure massage therapy may be mediated by increased vagal activity and gastric motility.

Exercise 2

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Paterson C, Allen JA, Browning M, Barlow G, Ewings P. A pilot study of therapeutic massage for people with Parkinson's disease: the added value of user involvement. Complement Ther Clin Pract. 2005 Aug;11(3):161-71. PMID: 16005833

OBJECTIVE: To carry out a pilot study, with particular attention to adequacy of outcome measures.

DESIGN: Observational study and user participation.

SETTING AND PARTICIPANTS: A local user group selected seven participants with a wide range of illness severity.

INTERVENTION: A course of eight 1hr sessions of deep whole body (therapeutic) massage over 8 weeks.

OUTCOME MEASURES: The Parkinson's Disease Questionnaire (PDQ-39), the Measure Yourself Medical Outcome Profile (MYMOP), and the Medication Change Questionnaire (MCQ). Semi-structured interviews, before and after the intervention.

RESULTS: In addition to enjoying the massage, individuals showed improvement in self-confidence, well-being, walking and activities of daily living. There was good agreement between data from the outcome questionnaires, interviews and clinical notes. User involvement highlighted issues that would otherwise have been ignored.

CONCLUSIONS: The study confirms the benefits of involving users in the research process and makes recommendations concerning the design of any future randomised trial.

Exercise 3

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Williams HL, Cullen LA, Barlow JH. The psychological well-being and self-efficacy of carers of children with disabilities following attendance on a simple massage training and support programme: a 12-month comparison study of adherers and non-adherers. Complement Ther Med. 2005 Jun;13(2):107-14. PMID: 16036168

OBJECTIVES: The Training and Support Programme (TSP) is an 8-week programme in which carers of children with disabilities receive instruction in simple massage techniques to use with their child. The aims of the present study were firstly to compare, adherers and non-adherers on measures of psychological well-being and self-efficacy and secondly, to examine whether, for adherers, the positive benefits of attending the TSP reported immediately after the Programme were maintained at 12-month follow-up.

DESIGN: Eighty-two carers took part in the study. Data were collected 12-months after completion of the TSP by self-report questionnaires mailed to carers. For comparisons between adherers and non-adherers at 12-month follow-up a between-groups design was used. For comparisons over time, a within-subjects design was used.

RESULTS: Adherers had significantly higher levels of self-efficacy for managing their child's psychosocial well-being, self-efficacy for carrying out the massage, and significantly better psychological well-being at follow-up compared to non-adherers. Furthermore, there were no significant differences over time on self-efficacy for managing their child's psychosocial well-being, self-efficacy for giving massage and levels of anxiety and depression at 12-month follow-up for adherers, suggesting that their improvements noted immediately post-programme were maintained at follow-up. Significant decreases on self-efficacy and depression were noted for non-adherers and there was a trend towards deterioration in anxious mood.

CONCLUSION: In conclusion, the present study suggests that the positive benefits the TSP has for carers of children with disabilities can be maintained if carers continue to practise the massage at home with their child.

Exercise 4

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Other times, the Abstract is pretty, well, abstract---it is a description of the article in fairly high-level narrative terms, rather than a concrete IMR(a)D description.

Let's look at some examples from the massage research literature:

Field T, Hernandez-Reif M, Diego M, Schanberg S, Kuhn C. Cortisol decreases and serotonin and dopamine increase following massage therapy. Int J Neurosci. 2005 Oct;115(10):1397-413. PMID: 16162447

In this article the positive effects of massage therapy on biochemistry are reviewed including decreased levels of cortisol and increased levels of serotonin and dopamine. The research reviewed includes studies on depression (including sex abuse and eating disorder studies), pain syndrome studies, research on auto-immune conditions (including asthma and chronic fatigue), immune studies (including HIV and breast cancer), and studies on the reduction of stress on the job, the stress of aging, and pregnancy stress. In studies in which cortisol was assayed either in saliva or in urine, significant decreases were noted in cortisol levels (averaging decreases 31%). In studies in which the activating neurotransmitters (serotonin and dopamine) were assayed in urine, an average increase of 28% was noted for serotonin and an average increase of 31% was noted for dopamine. These studies combined suggest the stress-alleviating effects (decreased cortisol) and the activating effects (increased serotonin and dopamine) of massage therapy on a variety of medical conditions and stressful experiences.

This abstract contains a very important lesson, one that gets right at the heart of evaluating the quality of research.

Over in the Journal Club, we've seen work by Christopher Moyer where he demonstrates that Tiffany Field's work in claiming to have shown significant decreases in cortisol contains serious methodological flaws, and, in fact, her conclusion is not backed up by the evidence.

In order to demonstrate that, he needed access to the entire article to see her methodology, and to the data to try to replicate the calculations.

That abstract does not provide sufficient detail to carry out that work.

What that means is that you cannot evaluate the quality of the article from the abstract alone--you simply do not have enough information available to you.

In order to read massage research and evaluate it adequately, you need to get the full article and read it carefully. Making decisions based only on the abstracts of articles means that you are putting your full trust in the authors, whether or not they made have made any errors in their work.

Exercise 5

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Agarwal A, Ranjan R, Dhiraaj S, Lakra A, Kumar M, Singh U. Acupressure for prevention of pre-operative anxiety: a prospective, randomised, placebo controlled study. Anaesthesia. 2005 Oct;60(10):978-81. PMID:16179042

Pre-operative anxiety is associated with many unwanted effects such as increased analgesic and anaesthetic requirement, postoperative pain and prolonged hospital stay. In the present study, we investigated the effects of acupressure on pre-operative anxiety and bispectral index (BIS) values. Seventy-six adults, ASA grade I and II, undergoing elective surgery, were randomly assigned to two equal groups. Group 1 (control) received acupressure at an inappropriate site and group 2 (acupressure) received acupressure at extra 1 point. The study was conducted during the pre-operative period and the duration of the study was 40 min (acupressure was applied for 10 min and thereafter patients were observed for another 30 min). Anxiety was recorded on a visual stress scale (VSS) at the start of the study and thereafter at 10 and 40 min. BIS was recorded at 0, 2, 5, 10, 12, 15, 30 and 40 min. The VSS decreased in both groups following pressure application for 10 min: median VSS (interquartile range) were 5 (1) vs. 8 (1) in the acupressure and 7 (0) vs. 8 (1) in the control groups (p < 0.001). Both pre-operative anxiety and BIS decreased significantly during acupressure application at extra 1 point (p < 0.001). Acupressure is effective in decreasing both pre-operative anxiety and BIS; however, these effects are not sustained 30 min following release of acupressure. Further studies are needed to elucidate the duration for which acupressure is effective.

Exercise 6

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Chen HM, Chang FY, Hsu CT. Effect of acupressure on nausea, vomiting, anxiety and pain among post-cesarean section women in Taiwan. Kaohsiung J Med Sci. 2005 Aug;21(8):341-50. PMID: 16158876

The purpose of this study was to examine the effectiveness of acupressure for controlling post-cesarean section (CS) symptoms, such as nausea and vomiting, anxiety perception and pain perception. A total of 104 eligible participants were recruited by convenience sampling of operating schedules at two hospitals. Participants assigned to the experimental group received acupressure, and those assigned to the control group received only postoperative nursing instruction. The experimental group received three acupressure treatments before CS and within the first 24 hours after CS. The first treatment was performed the night before CS, the second was performed 2-4 hours after CS, and the third was performed 8-10 hours after CS. The measures included the Rhodes Index of Nausea and Vomiting, Visual Analog Scale for Anxiety, State-Trait Anxiety Inventory, Visual Analog Scale for Pain, and physiologic indices. Statistical methods included percentages, mean value with standard deviation, t test and repeated measure ANOVA. The use of acupressure reduced the incidence of nausea, vomiting or retching from 69.3% to 53.9%, compared with control group (95% confidence interval = 1.65-0.11; p = 0.040) 2-4 hours after CS and from 36.2% to 15.4% compared with control group (95% confidence interval = 0.59-0.02; p = 0.024) 8-10 hours after CS. Results indicated that the experimental group had significantly lower anxiety and pain perception of cesarean experiences than the control group. Significant differences were found in all physiologic indices between the two groups. In conclusion, the utilization of acupressure treatment to promote the comfort of women during cesarean delivery is strongly recommended.

Exercise 7

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Iwasaki M. Interventional study on fatigue relief in mothers caring for hospitalized children--effect of massage incorporating techniques from Oriental medicine. Kurume Med J. 2005;52(1-2):19-27. PMID: 16119609 Free fulltext article

The study objective was to clarify the effect of massage on mothers caring for their hospitalized children. We conducted a comparative analysis of whether palm and shoulder massage could mitigate the physical and mental exhaustion experienced by such mothers. Subjects were 68 mothers whose children were admitted to the Department of Pediatrics, Kurume University Hospital with illnesses of varying severity. Twenty mothers living in Kurume City with healthy children were used as controls. A Japanese version of the Profile of Mood States (POMS) was employed as a mental index. Deep body temperature (frontal and palmar), systolic/diastolic blood pressure and heart rate were measured as physical indices before and after massage. The POMS scores for "Tension-Anxiety (T-A)", "Depression-Dejection (D)", "Anger-Hostility (A-H)", "Fatigue (F)" and "Confusion (C)" were significantly higher, and for "Vigor (V)" were significantly lower in mothers with hospitalized children than in the control group. Systolic blood pressures were also lower than those in the control group. After massage, T-A, D, A-H, F and C scores in the mothers with hospitalized children decreased and their V scores increased significantly. However, improvement in overall POMS scores was less than in the control group. And also improvement in each of the POMS scales was less than in the control group. Moreover, T-A scores in mothers of children with cancer were significantly higher than those in mothers of children suffering from other types of diseases. Our study demonstrated that mothers with hospitalized children were much more stressed than those with healthy children. The difference in the child's illness tended to exacerbate the degree of the mothers' mental fatigue. Massage has a favorable effect on stressed mothers and may be expected to serve as a useful supporting tool.

Exercise 8

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Mackereth P, Sylt P, Weinberg A, Campbell G. Chair massage for carers in an acute cancer hospital. Eur J Oncol Nurs. 2005 Jun;9(2):167-79. PMID: 15944109

The Chair Massage service considered in this evaluation study was provided to carers, visiting in-patients at a major cancer hospital in the UK. The two-stage evaluation comprised: firstly, a retrospective review of treatment records for the previous 12 months (n=182), and secondly, a prospective study, gathering data by interview and a 'next-day' questionnaire from carers (n=34), during 1 week of service delivery. The study at both stages sought to identify who used the service, post-treatment comments and changes in scores using a Feeling Good Thermometer (Field, T., 2000. Touch Therapy. Churchill Livingstone, London). During the second stage the carers were also asked about their concerns and worries, and to report changes in physical and emotional states using visual scales. Findings included significant improvements in physical and psychological scores; these were retained through to the next day. The next-day questionnaire also reported improved sleep for the majority of carers. A number of concerns and worries were raised at interview, notably anxieties about the patient and uncertainty about the future, family and financial worries. Overall, the service was well evaluated with parents and in particular female carers appearing to gain the most from the intervention.

Exercise 9

1. What research question was being investigated, and why?
2. What population was the treatment was studied in?
3. How was the treatment was studied (the methodology)?
4. What results did the investigator get for the outcome, including the statistical significance of those results?
5. What conclusions did the investigator draw about the study, including the meaning for clinical practice, and the basis for further research?

Both types of Abstract, however, serve the same purpose: they are intended to give you enough information to decide whether the article is relevant to the research you are interested in. If so, then that implies that taking the trouble to obtain and read the full articles is a good use of your time.

DO: Use the Abstract as it was intended---decide if an article interests you enough to bother getting it and reading it.

Example: Let's say I'm interested in exploring the safety of massage, because I am considering making a proposal to a hospital for massage for patients in the intensive care unit (ICU). I do a PubMed search on

"Massage"[MeSH] AND (safety OR risk)

This article is one of the ones returned from that search:

Ernst E. The safety of massage therapy. Rheumatology (Oxford). 2003 Sep;42(9):1101-6. Epub 2003 May 30. PMID: 12777645 Free full text

OBJECTIVES: After many years out of the limelight, massage therapy is now experiencing a revival. The aim of this systematic review is to evaluate its potential for harm.

METHODS: Computerized literature searches were carried out in four databases. All articles reporting adverse effects of any type of massage therapy were retrieved. Adverse effects relating to massage oil or ice were excluded. No language restrictions were applied. Data were extracted and evaluated according to predefined criteria.

RESULTS: Sixteen case reports of adverse effects and four case series were found. The majority of adverse effects were associated with exotic types of manual massage or massage delivered by laymen, while massage therapists were rarely implicated. The reported adverse events include cerebrovascular accidents, displacement of a ureteral stent, embolization of a kidney, haematoma, leg ulcers, nerve damage, posterior interosseous syndrome, pseudoaneurism, pulmonary embolism, ruptured uterus, strangulation of neck, thyrotoxicosis and various pain syndromes. In the majority of these instances, there can be little doubt about a cause-effect relationship. Serious adverse effects were associated mostly with massage techniques other than 'Swedish' massage.

CONCLUSION: Massage is not entirely risk free. However, serious adverse events are probably true rarities.

This article looks like it is very relevant to my proposal. I'll put it on my list to take to my librarian to get a hard copy of the article, or download it myself if I can.

Beachy JM. Premature infant massage in the NICU. Neonatal Netw. 2003 May-Jun;22(3):39-45. PMID: 12795507

Infant massage therapy is an inexpensive tool that should be utilized as part of the developmental care of the preterm infant. Nurses have been hesitant to begin massage therapy for fear of overstimulating the infant and because there has been insufficient research to prove its safety. Recent research, however, has shown that the significant benefits of infant massage therapy far outweigh the minimal risks. When infant massage therapy is properly applied to preterm infants, they respond with increased weight gains, improved developmental scores, and earlier discharge from the hospital. Parents of the preterm infant also benefit because infant massage enhances bonding with their child and increases confidence in their parenting skills. This article discusses the benefits and risks of massage for preterm infants and their families and explains how to implement massage therapy in the neonatal intensive care setting.

Although this article is about infants, and my proposal is for adults, nevertheless, it does deal with safety in the intensive-care unit, and so it is highly relevant to my proposal. I add it to my list to get at the library, as well.

What type of validity should I consider in this difference between populations in figuring out how this fits into my proposal?

Richards K, Nagel C, Markie M, Elwell J, Barone C. Use of complementary and alternative therapies to promote sleep in critically ill patients. Crit Care Nurs Clin North Am. 2003 Sep;15(3):329-40. PMID: 12943139

The efficacy of complementary and alternative therapies for sleep promotion in critically ill patients is largely unexamined. We found only seven studies (three on environmental interventions and one each on massage, music therapy, therapeutic touch, and, melatonin) that examined the effect of complementary and alternative therapies. A number of studies, however, have shown that massage, music therapy. and therapeutic touch promote relaxation and comfort in critically ill patients, which likely leads to improved sleep. Massage, music therapy, and therapeutic touch are safe for critically ill patients and should be routinely applied by ICU nurses who have received training on how to administer these specialized interventions. Environmental interventions, such as reducing noise, playing white noise such as ocean sounds, and decreasing interruptions to sleep for care, also are safe and logical interventions that ICU nurses should use to help patients sleep. Progressive muscle relaxation has been extensively studied and shown to be efficacious for improving sleep in persons with insomnia; however, progressive muscle relaxation requires that patients consciously attend to relaxing specific muscle groups and practice these techniques, which may be difficult for critically 11 patients. We do not currently recommend aromatherapy and alternative sedatives, such as valerian and melatonin, for sleep promotion in critically ill patients because the safety of these substances is unclear. In summary, we recommend that ICU nurses implement music therapy, environmental interventions, therapeutic touch, and relaxing massage to promote sleep in critically ill patients. These interventions are safe and may improve patient sleep, although randomized controlled trials are needed to test their efficacy. Aromatherapy and alternative sedatives require further investigation to determine their safety and efficacy.

This article is also very relevant, so it also gets added to my list.

Borruat FX, Kawasaki A. Optic nerve massaging: an extremely rare cause of self-inflicted blindness. Am J Ophthalmol. 2005 Apr;139(4):715-6. PMID: 15808173

PURPOSE: To report a patient whose self-inflicted blindness was not clinically apparent by history or external signs of trauma.

DESIGN: Observational case report.

METHODS: A 12-year-old girl with a history of social and behavioral problems was noted to have visual loss in her right eye. Examination revealed no light perception, optic nerve atrophy, partial upper lid ptosis, exotropia, and hypoesthesia of the cheek, all on her right side.

RESULTS: After undergoing extensive examinations which were unrevealing for a diagnosis, the patient admitted to a recurrent maneuver, which she secretly used to relieve anxiety and stress. This maneuver consisted of inserting her index finger under the right supraorbital rim and forcibly subluxating her globe out of the orbital space.

CONCLUSIONS: Self-inflicted visual loss can occur in nonpsychotic and nonviolent patients. Accurate diagnosis is important, as there is a risk of similar involvement to the fellow eye, and referral for psychiatric counseling is mandatory.

While this article comes up in a search on safety and massage, it is about self-injurious behavior by a young girl with adjustment problems. No massage practitioner would perform this particular massage, and there is no relevance to the intensive-care setting, either. For the purposes of this proposal, I would not consider this article relevant, and so I would not bother getting and reading it. The Abstract helped me save my time for relevant articles, by not wasting it on this irrelevant one.

Grant AC, Wang N. Carotid dissection associated with a handheld electric massager.  South Med J. 2004 Dec;97(12):1262-3. PMID: 15646768

The extracranial internal carotid artery (ICA) is susceptible to injury and dissection from external shear forces applied to the neck. Traumatic ICA dissection usually occurs in the setting of a sudden, high amplitude force causing significant distortion of surrounding soft tissues. Weaker, repetitive forces applied for longer intervals may also pose a risk for ICA dissection. A 38-year-old woman with no significant stroke risk factors had sudden onset of severe dysarthria and left hemiparesis several days after receiving an approximately 20-minute neck massage with a handheld electric massager. The moving elements consisted of two approximately 2-cm-diameter spheres that percuss the skin with low amplitude and high frequency. Magnetic resonance imaging and angiography demonstrated acute infarction in the right middle cerebral artery territory and dissection of the extracranial right ICA. Handheld electric massager units may cause ICA dissection and disabling stroke.

This article, as well, is not particularly relevant to my proposal---it deals with the risk from a massage device. It also does not have anything to do with intensive care, so it, as well, does not go on my list of articles to obtain and read.

These are examples of the appropriate use of Abstracts---to decide whether the article is interesting and relevant enough to you to proceed to bother with getting and reading the whole article.

DON'T: Use the information in the Abstract instead of taking the trouble to get the article.

This may seem like a good time-saving strategy, and it would be---if you could trust the Abstract to be correct. Unfortunately, much of the time you can't, and since---without checking it against the article---you can't know which Abstracts are accurate and which aren't.

Remember the old saying that when you "assume", you make an "ass'' out of "u" and "me"?

If the article author always did their job right, you could assume that the Abstract is always correct---but very often, the author does NOT do their job right.

It is very damaging, not to mention embarrassing, to pass on incorrect information, and there is far too much of that in the world already---we do not want to do more of it ourselves. Additionally, it is inconsistent with our maxim primum non nocere--"First, do no harm".

As much extra effort as it may seem to be, the confidence that you gain in the correctness of the information makes getting the article worthwhile.

Enough abstracts are missing important information, or they get important information wrong, that it's important and worthwhile to read the article--both to make sure that you get the correct information, and also to make sure that you have the entire context.

Pitkin RM, Branagan MA. Can the accuracy of abstracts be improved by providing specific instructions? A randomized controlled trial. JAMA. 1998 Jul 15;280(3):267-9. PMID: 9676677

CONTEXT: The most-read section of a research article is the abstract, and therefore it is especially important that the abstract be accurate.

...

MEAN OUTCOME MEASURE: Proportion of abstracts containing 1 or more of the following defects: inconsistency in data between abstract and body of manuscript (text, tables, and figures), data or other information given in abstract but not in body, and/or conclusions not justified by information in the abstract.

RESULTS: Of 250 manuscripts randomized, 13 were never revised and 34 were lost to follow-up, leaving a final comparison between 89 in the intervention group and 114 in the control group. Abstracts were defective in 25 (28%) and 30 (26%) cases, respectively (P=.78). Among 55 defective abstracts, 28 (51%) had inconsistencies, 16 (29%) contained data not present in the body, 8 (15%) had both types of defects, and 3 (5%) contained unjustified conclusions.

CONCLUSIONS: Defects in abstracts, particularly inconsistencies between abstract and body and the presentation of data in abstract but not in body, occur frequently. Specific instructions to authors who are revising their manuscripts are ineffective in lowering this rate. Journals should include in their editing processes specific and detailed attention to abstracts.

The errors in research article abstracts that the authors found include

1. what the abstract said was inconsistent with what the main body of the article said;
2. the abstract contained data that the main body of the article did not contain; and
3. the abstract contained conclusions that were not justified by the main body of the article.

Another article found additional problems in relying upon abstracts (MEDLINE, mentioned here, is part of PubMed):

[No authors listed]. Read MEDLINE abstracts with a pinch of salt. Lancet. 2006 Oct 21;368(9545):1394. PMID: 17055921

Read MEDLINE abstracts with a pinch of salt

With over 16 million citations and 7 million abstracts, the National Library of Medicine's MEDLINE has become an indispensable aid to research, diagnosis, and treatment. Health-care professionals, providers, and consumers around the world rely on this database as a trustworthy source of up-to-date information. Despite being designed as a retrieval tool, the fact is that for want of time, full-text access, or critical appraisal skills, many users depend solely on MEDLINE abstracts to inform decisions about care. Yet abstracts are known to be fickle representations of an article; it is the full text in which details of methods, funding, and conflicts of interest are found.

The inherent danger of abstracts was highlighted on Oct 2 by the Center for Science in the Public Interest, which exposed the inclusion in MEDLINE of a supplement to the Journal of the American College of Nutrition (JACN) devoted to down-playing the potential—albeit uncertain—health risks associated with current levels of salt intake. The guest editor was an adviser to the sodium committee of the International Life Sciences Institute, which funded the supplement, and whose corporate membership reads like a Who's Who of food manufacturing. Furthermore, several contributing authors can be found by searching the Salt Institute website, another industry-sponsored forum. In contradiction to JACN's stated standards, the supplement was not peer-reviewed, carried no conflict of interest statements, and failed to acknowledge funding on the title page of each review.

JACN is now strengthening its disclosure policy for supplements and the National Library of Medicine is revising indexing rules. From January, 2007, more rigorous standards will make it less likely that industry-sponsored supplements are indexed. But why tarnish MEDLINE's reputation by publishing them at all? Industry-sponsored supplements are merely the selling of a journal's brand name for advertising dressed as research. The resulting pseudo-science pollutes the literature and has no place on MEDLINE's publicly-funded website.

In addition to the issues listed above, abstracts don't contain sufficient information about details of methods, funding, or conflicts of interest as well.

These are all serious problems for evaluating research; they show that the abstract alone cannot be depended on for anything more than deciding whether or not to get and read the complete article.

Let's end with a teaching example that I use in my classes to illustrate the risks of relying solely on the abstract.

Here is the abstract of a case study on lymphedema with first onset 30 years after the removal of the patient's breast, due to cancer.

Brennan MJ, Weitz J. Lymphedema 30 years after radical mastectomy. Am J Phys Med Rehabil. 1992 Feb;71(1):12-4. PMID: 1739437

Lymphedema is a common and often distressing consequence of the management of breast cancer. This is the first report of new onset lymphedema 30 yr after radical mastectomy for the management of breast cancer. The patient's swelling began 10 days after commencing daily finger sticks to follow blood sugar, stemming from a recent diagnosis of diabetes mellitus. If proper precautions had been observed, swelling may not have developed. The patient had marked reduction of her edema with appropriate use of graded compression garments and self administered massage therapy, after early recognition and prompt intervention into her problem. This case highlights efforts that need to be made to prevent the development of edema.

And here is what the class and I discuss:

This article is a case study, interesting primarily because it reports a case of lymphedema occuring 30 years after radical mastectomy. According to the authors, there were no other similar cases in the literature. Lymphedema is a common result of surgery or radiation for the treatment of breast cancer, and it can have devastating consequences for the quality of the patient's life. However, the 30-year gap between surgery and the onset of lymphedema reported in this article appears to be unique in the literature.

The patient was a 78-year-old woman who had undergone a radical mastectomy of her left breast in 1960, and who had been regularly followed up with no signs of lymphedema until 1990, when she reported swelling in her left hand.

It developed that she had been diagnosed as diabetic 2 months previously, and was monitoring her blood sugar using daily fingersticks, as prescribed by her physician. As she was right-handed, she took the blood from the fingers of her left hand.

Apparently, the minor inflammation that resulted from the needle sticks, perhaps complicated by her age and diabetes, was enough to overwhelm her lymphatic system with additional fluid. Fortunately, she was diagnosed early and responded well to treatment.

Although this is an interesting case, which ended well, from our perspective there is not a lot in this article that adds to the body of knowledge about research on massage. The discussion of lymphedema may be useful and the article for practitioners who specialize in the area, but otherwise it probably is not necessary to take the trouble to seek out.

There is an illustration however, of the problem we discussed earlier of discrepancies between the abstract and the article.

The abstract states that "The patient had marked reduction of her edema with appropriate use of graded compression garments and self-administered massage therapy", yet there is no mention of massage therapy in the article itself at all.

A likely explanation is that the authors knew the patient was performing some kind of massage therapy, and remembered to mention it in the abstract, yet forgot to add the reference in the text. Or, perhaps, they were considering the compression garment as a kind of massage therapy; it is unclear from the text which interpretation is the right one.

In any case, this illustrates the risk of relying on abstracts-as we saw earlier, many abstracts can contain errors or discrepancies, when compared to the text. So reading only abstracts to evaluate the research literature is a risky strategy, as this example illustrates.

Next steps

The next chapter will wrap up all the work we did---in it, we'll discuss how we use the information in this book, and put it into practice on a real article from the literature.