Michael Gellis compares liposuction satisfaction rates between low and high body mass index patients under awake local tumescent anesthesia in 1,826 cases.
Since liposuction began in the United States in the early 1980s, plastic surgeons have always been concerned with patient satisfaction1. Somehow, even with the use of large cannulas, now considered archaic, the results seemed to be reasonable in our biased opinion. We thought and preached, often statistically undocumented, that patients were generally satisfied. During training, our chief Robert Pool, MD, often stated, ‘If a patient tells you she is pleased with her results, she means she likes you.’
There were also groups of patients plastic surgeons were told to avoid performing liposuction upon: the very low body mass index (BMI) and high BMI patients. The reason given, also without statistical support, was that these patients would never be satisfied. This begs the question, considering liposuction, whether or not there is a difference in the rates of satisfaction among patients with different BMIs?
In 1997, the World Health Organization (WHO) introduced today’s BMI standards, recognizing a BMI of 30 or more as obese. Soon after, in 1998, the National Heart, Lung, and Blood Institute (NHLBI) and National Institutes of Health (NIH) published the same criteria2. Therefore, in this study, a BMI equal to or greater than 30 was considered obese, as opposed to a BMI less than 30 being considered normal or simply overweight.
Unmonitored self-assessment surveys have been consistently used to evaluate satisfaction and obtain data from patients since there were no suggestions from a medical provider to influence the answer. The patient could give accurate information as to what they believed to be factual of their experience and results. Most unbiased surveys were either conducted by mail or internet to provide identification privacy.
This retrospective study was conducted to compare post-operative liposuction satisfaction rates between low and high BMI patients ranging from 18 to 43 BMI.
A 3-month post-operative survey was collected from 1,826 patients, during the time period between 1/9/13 through 30/8/14. A total of 9,540 surveys were sent out and 1,826 were returned, giving a response rate of 19.14%. The survey was conducted online via email through a survey provider (SurveyGizmo, Boulder, CO).
When the surveys were returned, the data was divided according to BMIs and satisfaction scores where a BMI of 29 and below was considered normal or overweight, and a BMI of 30 or above was considered obese. Associated with those divisions, the number of patients and the average satisfaction score for each BMI category were recorded. We also recorded the patient’s age, but did not consider satisfaction scores compared to age in this study.
In addition to questions about the patient’s surgery, the survey also asked a series of questions, not necessarily related to the procedure, in order to improve our own practice’s system and office procedures. These questions, answered on a scale of 1 to 5, are listed in Table 1. The actual evaluation of patients’ satisfaction, however, was just one of the questions (Question 6), buried within the 21-question survey: ‘How well did we meet your expectations?’
After the data was collected, patients were divided into two categories: high (≥30) and low (<30) BMI; the categories contained 740 and 1,086 observations respectively. The average satisfaction scores, numbers of patients within each BMI category, and percentage of total are shown in Table 2, this data is plotted in Figure 1.
A univariate Ordinary Least Squares (OLS) regression was used to compare the impact of patient BMI on satisfaction. Ordinary Least Squares regression is a commonly used statistical technique in research, featuring predominantly in biostatistics, engineering, finance, and economic applications. It allows researchers to account for the variation in a dependent variable (in this case satisfaction on a 5 point scale) given one or more independent variables, such as a binary variable for high or low BMI. In this case, satisfaction values were regressed on a dummy variable, ‘Low’. The resulting estimation is:
Satisfaction = 3.72 – .053 *Low
This indicates that the average satisfaction for a ‘high’ BMI patient is 3.72, and the average satisfaction for a ‘low’ BMI patient is actually lower: 3.667. This regression holds very low predictive power, predicting only 0.047% of the variation in satisfaction between patients, indicating that virtually all of the variation (the other 99.953%) not predicted by the model is explained by other factors, such as a surgeon’s technical ability or bedside manner.
In testing whether Low is not distinct from zero, the null and alternative hypotheses are:
H0: β1 = 0
Ha: β1 ≠ 0
The null of a two sided t-test is rejected at the 5% level if the absolute value of the t statistic is greater than 1.96. The t statistic of β1 is -.92; consequently the null hypothesis cannot be rejected. The difference in satisfaction between patients cannot be shown to be statistically distinct from zero with a reasonable degree of certainty.
Comments concerning liposuction satisfaction were first made by Carson Lewis, MD, in 19841. Dr. Lewis emphasized that the ultimate goal was patient satisfaction, which could be achieved by selecting patients with realistic expectations and positive attitudes, who were highly likely to be satisfied regardless of the outcome. He suggested that the optimal patient should:
- Be self-motivated and not motivated by an outside party
- Have had the concern for a long time and should not treat it lightly
- Have a higher level of deformity than anxiety
- Not be looking for a complete makeover
- Not be severely depressed or psychotic.
Frank Herhahn, MD, also offered guidelines for choosing the ideal patient: ‘usually less than 35 years old and falls within the ideal weight for his or her height and body type,’ although the source listed no method, such as BMI, for determining the ‘ideal weight’ for height and body type3. The patient should also have good skin turgor, no cellulite, firm adiposity, little sagging, and no striae. The so-called ‘average’ patient at that time was 30 to 45 years old, and 5 to 10 pounds above the ideal weight ‘for his or her height and body build.’ At that time, the ‘less-than-ideal patient’ was over 35 years old and 10 to 20 pounds overweight. Those comments were recorded in 1984 when liposuction began in the United States. In today’s world, those ideal patients only represent a tiny minority of patients upon whom plastic surgeons actually perform liposuction.
There has been minimal discussion concerning satisfaction rates in high BMI liposuction patients. Within the literature, the following has been researched. When patients’ BMIs were included in an article about satisfaction after liposuction and abdominoplasty, 79.4% of those patients had a BMI less than 304. In that study, surveys were conducted by in-person interviews of 360 patients, and the results demonstrated an 82.5% satisfaction rate ‘of liposuction patients whose results met or exceeded their expectations.’ That study compared its results to other studies which documented satisfaction rates ranging from 76 to 85%. However, one must realize that this data collecting was done through face-to-face biased interviewers, where patients were possibly influenced in giving their opinions.
On a different note, in a Continuing Medical Education (CME) article on liposuction in Plastic & Reconstructive Surgery in 2008, Ronald Iverson, MD, remarking on the topic of patient satisfaction, stated, ‘patients who were committed to a positive lifestyle change or who were already cognizant and practicing a “healthy lifestyle”…’ showed the highest satisfaction rates5.
Another study considering post-operative satisfaction between patients who gained weight or did not gain weight after liposuction was performed by Rod Rohrich, MD, in 2004 via a mailed questionnaire6, and concluded that regardless of whether patients gained or lost weight after liposuction, 73% of those who gained weight and 82% of those who did not were either very satisfied or satisfied with the procedure and that given the chance, 72% who gained weight and 82% who did not said they would have it done again. There was no mention, however, comparing satisfaction rates considering actual BMIs.
An article assessing the improvement in the quality of life and patient satisfaction following body contouring surgery was published in 2013 by Shehab Jabir7. Predominantly a psychological study, this research used an unbiased method to determine, among other things, patient satisfaction following body contouring surgery. Multiple questions were given to the patient without physician or technician bias added. The types of surgeries performed were mostly body contouring, such as lipectomy after massive weight loss. The questions, called ‘instruments’ for determining satisfaction and reaching expectations, determined that 61.5% of the results were in accordance with the patients’ expectations.
Other studies sought to compare the percentage of patients among different levels of satisfaction, for example comparing patients with different satisfaction ratings of ‘bad’, ‘okay’, ‘good’, or ‘very good’. Our study, however, averaged the satisfaction scores for each BMI in order to prove there is no statistical difference in these rates among high and low BMI categories, but did not include an analysis of the percentage of patients who rated their satisfaction scores as 1, 2, 3, 4, or 5. This comparison proved that there was no statistical difference in satisfaction among all BMIs.
There are other factors that this study did not evaluate and may have some bearing on results. For example, does age have a bearing on satisfaction rates? Or does percentage of fat removed have a similar bearing on satisfaction rates? Further studies will be needed to elucidate that.
Although conventional wisdom holds that high BMI patients would be less satisfied with liposuction than lower BMI patients, the findings in this sample refute this assertion. In fact, when considering BMI categories ranging from 18 to 43, the satisfaction rates among these groups were not statistically different. It is possible that a more evenly distributed sample would yield different results. The study was performed using an unbiased electronic survey which was more accurate than a face-to-face biased interviewing process. How age, weight gain, diet, and exercise affect post-operative satisfaction are other factors that need to be researched.